fix: eslint run failed

Co-authored-by: crazywoola <427733928@qq.com>

api/Dockerfile

@@ -52,12 +52,14 @@ RUN apt-get update \
     && apt-get install -y --no-install-recommends curl nodejs libgmp-dev libmpfr-dev libmpc-dev \
     # if you located in China, you can use aliyun mirror to speed up
     # && echo "deb http://mirrors.aliyun.com/debian testing main" > /etc/apt/sources.list \
-    && echo "deb http://deb.debian.org/debian testing main" > /etc/apt/sources.list \
+    && echo "deb http://deb.debian.org/debian bookworm main" > /etc/apt/sources.list \
     && apt-get update \
     # For Security
-    && apt-get install -y --no-install-recommends expat=2.6.4-1 libldap-2.5-0=2.5.19+dfsg-1 perl=5.40.0-8 libsqlite3-0=3.46.1-1 zlib1g=1:1.3.dfsg+really1.3.1-1+b1 \
+    && apt-get install -y --no-install-recommends expat libldap-2.5-0 perl libsqlite3-0 zlib1g \
     # install a chinese font to support the use of tools like matplotlib
     && apt-get install -y fonts-noto-cjk \
+    # install libmagic to support the use of python-magic guess MIMETYPE
+    && apt-get install -y libmagic1 \
     && apt-get autoremove -y \
     && rm -rf /var/lib/apt/lists/*
 

api/controllers/common/helpers.py

@@ -1,12 +1,32 @@
 import mimetypes
 import os
+import platform
 import re
 import urllib.parse
+import warnings
 from collections.abc import Mapping
 from typing import Any
 from uuid import uuid4
 
 import httpx
+
+try:
+    import magic
+except ImportError:
+    if platform.system() == "Windows":
+        warnings.warn(
+            "To use python-magic guess MIMETYPE, you need to run `pip install python-magic-bin`", stacklevel=2
+        )
+    elif platform.system() == "Darwin":
+        warnings.warn("To use python-magic guess MIMETYPE, you need to run `brew install libmagic`", stacklevel=2)
+    elif platform.system() == "Linux":
+        warnings.warn(
+            "To use python-magic guess MIMETYPE, you need to run `sudo apt-get install libmagic1`", stacklevel=2
+        )
+    else:
+        warnings.warn("To use python-magic guess MIMETYPE, you need to install `libmagic`", stacklevel=2)
+    magic = None  # type: ignore
+
 from pydantic import BaseModel
 
 from configs import dify_config
@@ -47,6 +67,13 @@ def guess_file_info_from_response(response: httpx.Response):
         # If guessing fails, use Content-Type from response headers
         mimetype = response.headers.get("Content-Type", "application/octet-stream")
 
+    # Use python-magic to guess MIME type if still unknown or generic
+    if mimetype == "application/octet-stream" and magic is not None:
+        try:
+            mimetype = magic.from_buffer(response.content[:1024], mime=True)
+        except magic.MagicException:
+            pass
+
     extension = os.path.splitext(filename)[1]
 
     # Ensure filename has an extension

api/controllers/console/datasets/datasets.py

@@ -620,7 +620,6 @@ class DatasetRetrievalSettingApi(Resource):
         match vector_type:
             case (
                 VectorType.RELYT
-                | VectorType.PGVECTOR
                 | VectorType.TIDB_VECTOR
                 | VectorType.CHROMA
                 | VectorType.TENCENT

api/controllers/console/explore/message.py

@@ -50,7 +50,7 @@ class MessageListApi(InstalledAppResource):
 
         try:
             return MessageService.pagination_by_first_id(
-                app_model, current_user, args["conversation_id"], args["first_id"], args["limit"], "desc"
+                app_model, current_user, args["conversation_id"], args["first_id"], args["limit"]
             )
         except services.errors.conversation.ConversationNotExistsError:
             raise NotFound("Conversation Not Exists.")

api/controllers/inner_api/workspace/workspace.py

@@ -1,3 +1,5 @@
+import json
+
 from flask_restful import Resource, reqparse  # type: ignore
 
 from controllers.console.wraps import setup_required
@@ -29,4 +31,34 @@ class EnterpriseWorkspace(Resource):
         return {"message": "enterprise workspace created."}
 
 
+class EnterpriseWorkspaceNoOwnerEmail(Resource):
+    @setup_required
+    @inner_api_only
+    def post(self):
+        parser = reqparse.RequestParser()
+        parser.add_argument("name", type=str, required=True, location="json")
+        args = parser.parse_args()
+
+        tenant = TenantService.create_tenant(args["name"], is_from_dashboard=True)
+
+        tenant_was_created.send(tenant)
+
+        resp = {
+            "id": tenant.id,
+            "name": tenant.name,
+            "encrypt_public_key": tenant.encrypt_public_key,
+            "plan": tenant.plan,
+            "status": tenant.status,
+            "custom_config": json.loads(tenant.custom_config) if tenant.custom_config else {},
+            "created_at": tenant.created_at.isoformat() if tenant.created_at else None,
+            "updated_at": tenant.updated_at.isoformat() if tenant.updated_at else None,
+        }
+
+        return {
+            "message": "enterprise workspace created.",
+            "tenant": resp,
+        }
+
+
 api.add_resource(EnterpriseWorkspace, "/enterprise/workspace")
+api.add_resource(EnterpriseWorkspaceNoOwnerEmail, "/enterprise/workspace/ownerless")

api/controllers/service_api/dataset/document.py

@@ -18,6 +18,7 @@ from controllers.service_api.app.error import (
 from controllers.service_api.dataset.error import (
     ArchivedDocumentImmutableError,
     DocumentIndexingError,
+    InvalidMetadataError,
 )
 from controllers.service_api.wraps import DatasetApiResource, cloud_edition_billing_resource_check
 from core.errors.error import ProviderTokenNotInitError
@@ -50,6 +51,9 @@ class DocumentAddByTextApi(DatasetApiResource):
             "indexing_technique", type=str, choices=Dataset.INDEXING_TECHNIQUE_LIST, nullable=False, location="json"
         )
         parser.add_argument("retrieval_model", type=dict, required=False, nullable=False, location="json")
+        parser.add_argument("doc_type", type=str, required=False, nullable=True, location="json")
+        parser.add_argument("doc_metadata", type=dict, required=False, nullable=True, location="json")
+
         args = parser.parse_args()
         dataset_id = str(dataset_id)
         tenant_id = str(tenant_id)
@@ -61,6 +65,28 @@ class DocumentAddByTextApi(DatasetApiResource):
         if not dataset.indexing_technique and not args["indexing_technique"]:
             raise ValueError("indexing_technique is required.")
 
+        # Validate metadata if provided
+        if args.get("doc_type") or args.get("doc_metadata"):
+            if not args.get("doc_type") or not args.get("doc_metadata"):
+                raise InvalidMetadataError("Both doc_type and doc_metadata must be provided when adding metadata")
+
+            if args["doc_type"] not in DocumentService.DOCUMENT_METADATA_SCHEMA:
+                raise InvalidMetadataError(
+                    "Invalid doc_type. Must be one of: " + ", ".join(DocumentService.DOCUMENT_METADATA_SCHEMA.keys())
+                )
+
+            if not isinstance(args["doc_metadata"], dict):
+                raise InvalidMetadataError("doc_metadata must be a dictionary")
+
+            # Validate metadata schema based on doc_type
+            if args["doc_type"] != "others":
+                metadata_schema = DocumentService.DOCUMENT_METADATA_SCHEMA[args["doc_type"]]
+                for key, value in args["doc_metadata"].items():
+                    if key in metadata_schema and not isinstance(value, metadata_schema[key]):
+                        raise InvalidMetadataError(f"Invalid type for metadata field {key}")
+            # set to MetaDataConfig
+            args["metadata"] = {"doc_type": args["doc_type"], "doc_metadata": args["doc_metadata"]}
+
         text = args.get("text")
         name = args.get("name")
         if text is None or name is None:
@@ -107,6 +133,8 @@ class DocumentUpdateByTextApi(DatasetApiResource):
             "doc_language", type=str, default="English", required=False, nullable=False, location="json"
         )
         parser.add_argument("retrieval_model", type=dict, required=False, nullable=False, location="json")
+        parser.add_argument("doc_type", type=str, required=False, nullable=True, location="json")
+        parser.add_argument("doc_metadata", type=dict, required=False, nullable=True, location="json")
         args = parser.parse_args()
         dataset_id = str(dataset_id)
         tenant_id = str(tenant_id)
@@ -115,6 +143,32 @@ class DocumentUpdateByTextApi(DatasetApiResource):
         if not dataset:
             raise ValueError("Dataset is not exist.")
 
+        # indexing_technique is already set in dataset since this is an update
+        args["indexing_technique"] = dataset.indexing_technique
+
+        # Validate metadata if provided
+        if args.get("doc_type") or args.get("doc_metadata"):
+            if not args.get("doc_type") or not args.get("doc_metadata"):
+                raise InvalidMetadataError("Both doc_type and doc_metadata must be provided when adding metadata")
+
+            if args["doc_type"] not in DocumentService.DOCUMENT_METADATA_SCHEMA:
+                raise InvalidMetadataError(
+                    "Invalid doc_type. Must be one of: " + ", ".join(DocumentService.DOCUMENT_METADATA_SCHEMA.keys())
+                )
+
+            if not isinstance(args["doc_metadata"], dict):
+                raise InvalidMetadataError("doc_metadata must be a dictionary")
+
+            # Validate metadata schema based on doc_type
+            if args["doc_type"] != "others":
+                metadata_schema = DocumentService.DOCUMENT_METADATA_SCHEMA[args["doc_type"]]
+                for key, value in args["doc_metadata"].items():
+                    if key in metadata_schema and not isinstance(value, metadata_schema[key]):
+                        raise InvalidMetadataError(f"Invalid type for metadata field {key}")
+
+            # set to MetaDataConfig
+            args["metadata"] = {"doc_type": args["doc_type"], "doc_metadata": args["doc_metadata"]}
+
         if args["text"]:
             text = args.get("text")
             name = args.get("name")
@@ -161,6 +215,30 @@ class DocumentAddByFileApi(DatasetApiResource):
             args["doc_form"] = "text_model"
         if "doc_language" not in args:
             args["doc_language"] = "English"
+
+        # Validate metadata if provided
+        if args.get("doc_type") or args.get("doc_metadata"):
+            if not args.get("doc_type") or not args.get("doc_metadata"):
+                raise InvalidMetadataError("Both doc_type and doc_metadata must be provided when adding metadata")
+
+            if args["doc_type"] not in DocumentService.DOCUMENT_METADATA_SCHEMA:
+                raise InvalidMetadataError(
+                    "Invalid doc_type. Must be one of: " + ", ".join(DocumentService.DOCUMENT_METADATA_SCHEMA.keys())
+                )
+
+            if not isinstance(args["doc_metadata"], dict):
+                raise InvalidMetadataError("doc_metadata must be a dictionary")
+
+            # Validate metadata schema based on doc_type
+            if args["doc_type"] != "others":
+                metadata_schema = DocumentService.DOCUMENT_METADATA_SCHEMA[args["doc_type"]]
+                for key, value in args["doc_metadata"].items():
+                    if key in metadata_schema and not isinstance(value, metadata_schema[key]):
+                        raise InvalidMetadataError(f"Invalid type for metadata field {key}")
+
+            # set to MetaDataConfig
+            args["metadata"] = {"doc_type": args["doc_type"], "doc_metadata": args["doc_metadata"]}
+
         # get dataset info
         dataset_id = str(dataset_id)
         tenant_id = str(tenant_id)
@@ -228,6 +306,29 @@ class DocumentUpdateByFileApi(DatasetApiResource):
         if "doc_language" not in args:
             args["doc_language"] = "English"
 
+        # Validate metadata if provided
+        if args.get("doc_type") or args.get("doc_metadata"):
+            if not args.get("doc_type") or not args.get("doc_metadata"):
+                raise InvalidMetadataError("Both doc_type and doc_metadata must be provided when adding metadata")
+
+            if args["doc_type"] not in DocumentService.DOCUMENT_METADATA_SCHEMA:
+                raise InvalidMetadataError(
+                    "Invalid doc_type. Must be one of: " + ", ".join(DocumentService.DOCUMENT_METADATA_SCHEMA.keys())
+                )
+
+            if not isinstance(args["doc_metadata"], dict):
+                raise InvalidMetadataError("doc_metadata must be a dictionary")
+
+            # Validate metadata schema based on doc_type
+            if args["doc_type"] != "others":
+                metadata_schema = DocumentService.DOCUMENT_METADATA_SCHEMA[args["doc_type"]]
+                for key, value in args["doc_metadata"].items():
+                    if key in metadata_schema and not isinstance(value, metadata_schema[key]):
+                        raise InvalidMetadataError(f"Invalid type for metadata field {key}")
+
+            # set to MetaDataConfig
+            args["metadata"] = {"doc_type": args["doc_type"], "doc_metadata": args["doc_metadata"]}
+
         # get dataset info
         dataset_id = str(dataset_id)
         tenant_id = str(tenant_id)

api/controllers/web/message.py

@@ -91,7 +91,7 @@ class MessageListApi(WebApiResource):
 
         try:
             return MessageService.pagination_by_first_id(
-                app_model, end_user, args["conversation_id"], args["first_id"], args["limit"], "desc"
+                app_model, end_user, args["conversation_id"], args["first_id"], args["limit"]
             )
         except services.errors.conversation.ConversationNotExistsError:
             raise NotFound("Conversation Not Exists.")

api/core/app/apps/agent_chat/app_runner.py

@@ -202,7 +202,7 @@ class AgentChatAppRunner(AppRunner):
         # change function call strategy based on LLM model
         llm_model = cast(LargeLanguageModel, model_instance.model_type_instance)
         model_schema = llm_model.get_model_schema(model_instance.model, model_instance.credentials)
-        if not model_schema or not model_schema.features:
+        if not model_schema:
             raise ValueError("Model schema not found")
 
         if {ModelFeature.MULTI_TOOL_CALL, ModelFeature.TOOL_CALL}.intersection(model_schema.features or []):

api/core/model_runtime/model_providers/__base/ai_model.py

@@ -221,13 +221,12 @@ class AIModel(ABC):
         :param credentials: model credentials
         :return: model schema
         """
-        # get predefined models (predefined_models)
-        models = self.predefined_models()
-
-        model_map = {model.model: model for model in models}
-        if model in model_map:
-            return model_map[model]
+        # Try to get model schema from predefined models
+        for predefined_model in self.predefined_models():
+            if model == predefined_model.model:
+                return predefined_model
 
+        # Try to get model schema from credentials
         if credentials:
             model_schema = self.get_customizable_model_schema_from_credentials(model, credentials)
             if model_schema:

api/core/model_runtime/model_providers/azure_openai/azure_openai.yaml

@@ -53,6 +53,9 @@ model_credential_schema:
       type: select
       required: true
       options:
+        - label:
+            en_US: 2024-12-01-preview
+          value: 2024-12-01-preview
         - label:
             en_US: 2024-10-01-preview
           value: 2024-10-01-preview

api/core/model_runtime/model_providers/bedrock/bedrock.yaml

@@ -44,6 +44,7 @@ provider_credential_schema:
       label:
         en_US: AWS Region
         zh_Hans: AWS 地区
+        ja_JP: AWS リージョン
       type: select
       default: us-east-1
       options:
@@ -51,62 +52,77 @@ provider_credential_schema:
           label:
             en_US: US East (N. Virginia)
             zh_Hans: 美国东部 (弗吉尼亚北部)
+            ja_JP: 米国 (バージニア北部)
         - value: us-east-2
           label:
             en_US: US East (Ohio)
-            zh_Hans: 美国东部 (弗吉尼亚北部)
+            zh_Hans: 美国东部 (俄亥俄)
+            ja_JP: 米国 (オハイオ)
         - value: us-west-2
           label:
             en_US: US West (Oregon)
             zh_Hans: 美国西部 (俄勒冈州)
+            ja_JP: 米国 (オレゴン)
         - value: ap-south-1
           label:
             en_US: Asia Pacific (Mumbai)
             zh_Hans: 亚太地区（孟买）
+            ja_JP: アジアパシフィック (ムンバイ)
         - value: ap-southeast-1
           label:
             en_US: Asia Pacific (Singapore)
             zh_Hans: 亚太地区 (新加坡)
+            ja_JP: アジアパシフィック (シンガポール)
         - value: ap-southeast-2
           label:
             en_US: Asia Pacific (Sydney)
             zh_Hans: 亚太地区 (悉尼)
+            ja_JP: アジアパシフィック (シドニー)
         - value: ap-northeast-1
           label:
             en_US: Asia Pacific (Tokyo)
             zh_Hans: 亚太地区 (东京)
+            ja_JP: アジアパシフィック (東京)
         - value: ap-northeast-2
           label:
             en_US: Asia Pacific (Seoul)
             zh_Hans: 亚太地区（首尔）
+            ja_JP: アジアパシフィック (ソウル)
         - value: ca-central-1
           label:
             en_US: Canada (Central)
             zh_Hans: 加拿大（中部）
+            ja_JP: カナダ (中部)
         - value: eu-central-1
           label:
             en_US: Europe (Frankfurt)
             zh_Hans: 欧洲 (法兰克福)
+            ja_JP: 欧州 (フランクフルト)
         - value: eu-west-1
           label:
             en_US: Europe (Ireland)
             zh_Hans: 欧洲（爱尔兰）
+            ja_JP: 欧州 (アイルランド)
         - value: eu-west-2
           label:
             en_US: Europe (London)
             zh_Hans: 欧洲西部 (伦敦)
+            ja_JP: 欧州 (ロンドン)
         - value: eu-west-3
           label:
             en_US: Europe (Paris)
             zh_Hans: 欧洲（巴黎）
+            ja_JP: 欧州 (パリ)
         - value: sa-east-1
           label:
             en_US: South America (São Paulo)
             zh_Hans: 南美洲（圣保罗）
+            ja_JP: 南米 (サンパウロ)
         - value: us-gov-west-1
           label:
             en_US: AWS GovCloud (US-West)
             zh_Hans: AWS GovCloud (US-West)
+            ja_JP: AWS GovCloud (米国西部)
     - variable: model_for_validation
       required: false
       label:

api/core/model_runtime/model_providers/cohere/llm/llm.py

@@ -677,16 +677,17 @@ class CohereLargeLanguageModel(LargeLanguageModel):
 
         :return: model schema
         """
-        # get model schema
-        models = self.predefined_models()
-        model_map = {model.model: model for model in models}
-
         mode = credentials.get("mode")
+        base_model_schema = None
+        for predefined_model in self.predefined_models():
+            if (
+                mode == "chat" and predefined_model.model == "command-light-chat"
+            ) or predefined_model.model == "command-light":
+                base_model_schema = predefined_model
+                break
 
-        if mode == "chat":
-            base_model_schema = model_map["command-light-chat"]
-        else:
-            base_model_schema = model_map["command-light"]
+        if not base_model_schema:
+            raise ValueError("Model not found")
 
         base_model_schema = cast(AIModelEntity, base_model_schema)
 

api/core/model_runtime/model_providers/deepseek/llm/llm.py

@@ -1,10 +1,13 @@
+import json
 from collections.abc import Generator
 from typing import Optional, Union
 
+import requests
 from yarl import URL
 
-from core.model_runtime.entities.llm_entities import LLMMode, LLMResult
+from core.model_runtime.entities.llm_entities import LLMMode, LLMResult, LLMResultChunk, LLMResultChunkDelta
 from core.model_runtime.entities.message_entities import (
+    AssistantPromptMessage,
     PromptMessage,
     PromptMessageTool,
 )
@@ -36,3 +39,208 @@ class DeepseekLargeLanguageModel(OAIAPICompatLargeLanguageModel):
         credentials["mode"] = LLMMode.CHAT.value
         credentials["function_calling_type"] = "tool_call"
         credentials["stream_function_calling"] = "support"
+
+    def _handle_generate_stream_response(
+        self, model: str, credentials: dict, response: requests.Response, prompt_messages: list[PromptMessage]
+    ) -> Generator:
+        """
+        Handle llm stream response
+
+        :param model: model name
+        :param credentials: model credentials
+        :param response: streamed response
+        :param prompt_messages: prompt messages
+        :return: llm response chunk generator
+        """
+        full_assistant_content = ""
+        chunk_index = 0
+        is_reasoning_started = False  # Add flag to track reasoning state
+
+        def create_final_llm_result_chunk(
+            id: Optional[str], index: int, message: AssistantPromptMessage, finish_reason: str, usage: dict
+        ) -> LLMResultChunk:
+            # calculate num tokens
+            prompt_tokens = usage and usage.get("prompt_tokens")
+            if prompt_tokens is None:
+                prompt_tokens = self._num_tokens_from_string(model, prompt_messages[0].content)
+            completion_tokens = usage and usage.get("completion_tokens")
+            if completion_tokens is None:
+                completion_tokens = self._num_tokens_from_string(model, full_assistant_content)
+
+            # transform usage
+            usage = self._calc_response_usage(model, credentials, prompt_tokens, completion_tokens)
+
+            return LLMResultChunk(
+                id=id,
+                model=model,
+                prompt_messages=prompt_messages,
+                delta=LLMResultChunkDelta(index=index, message=message, finish_reason=finish_reason, usage=usage),
+            )
+
+        # delimiter for stream response, need unicode_escape
+        import codecs
+
+        delimiter = credentials.get("stream_mode_delimiter", "\n\n")
+        delimiter = codecs.decode(delimiter, "unicode_escape")
+
+        tools_calls: list[AssistantPromptMessage.ToolCall] = []
+
+        def increase_tool_call(new_tool_calls: list[AssistantPromptMessage.ToolCall]):
+            def get_tool_call(tool_call_id: str):
+                if not tool_call_id:
+                    return tools_calls[-1]
+
+                tool_call = next((tool_call for tool_call in tools_calls if tool_call.id == tool_call_id), None)
+                if tool_call is None:
+                    tool_call = AssistantPromptMessage.ToolCall(
+                        id=tool_call_id,
+                        type="function",
+                        function=AssistantPromptMessage.ToolCall.ToolCallFunction(name="", arguments=""),
+                    )
+                    tools_calls.append(tool_call)
+
+                return tool_call
+
+            for new_tool_call in new_tool_calls:
+                # get tool call
+                tool_call = get_tool_call(new_tool_call.function.name)
+                # update tool call
+                if new_tool_call.id:
+                    tool_call.id = new_tool_call.id
+                if new_tool_call.type:
+                    tool_call.type = new_tool_call.type
+                if new_tool_call.function.name:
+                    tool_call.function.name = new_tool_call.function.name
+                if new_tool_call.function.arguments:
+                    tool_call.function.arguments += new_tool_call.function.arguments
+
+        finish_reason = None  # The default value of finish_reason is None
+        message_id, usage = None, None
+        for chunk in response.iter_lines(decode_unicode=True, delimiter=delimiter):
+            chunk = chunk.strip()
+            if chunk:
+                # ignore sse comments
+                if chunk.startswith(":"):
+                    continue
+                decoded_chunk = chunk.strip().removeprefix("data:").lstrip()
+                if decoded_chunk == "[DONE]":  # Some provider returns "data: [DONE]"
+                    continue
+
+                try:
+                    chunk_json: dict = json.loads(decoded_chunk)
+                # stream ended
+                except json.JSONDecodeError as e:
+                    yield create_final_llm_result_chunk(
+                        id=message_id,
+                        index=chunk_index + 1,
+                        message=AssistantPromptMessage(content=""),
+                        finish_reason="Non-JSON encountered.",
+                        usage=usage,
+                    )
+                    break
+                # handle the error here. for issue #11629
+                if chunk_json.get("error") and chunk_json.get("choices") is None:
+                    raise ValueError(chunk_json.get("error"))
+
+                if chunk_json:
+                    if u := chunk_json.get("usage"):
+                        usage = u
+                if not chunk_json or len(chunk_json["choices"]) == 0:
+                    continue
+
+                choice = chunk_json["choices"][0]
+                finish_reason = chunk_json["choices"][0].get("finish_reason")
+                message_id = chunk_json.get("id")
+                chunk_index += 1
+
+                if "delta" in choice:
+                    delta = choice["delta"]
+                    is_reasoning = delta.get("reasoning_content")
+                    delta_content = delta.get("content") or delta.get("reasoning_content")
+
+                    assistant_message_tool_calls = None
+
+                    if "tool_calls" in delta and credentials.get("function_calling_type", "no_call") == "tool_call":
+                        assistant_message_tool_calls = delta.get("tool_calls", None)
+                    elif (
+                        "function_call" in delta
+                        and credentials.get("function_calling_type", "no_call") == "function_call"
+                    ):
+                        assistant_message_tool_calls = [
+                            {"id": "tool_call_id", "type": "function", "function": delta.get("function_call", {})}
+                        ]
+
+                    # assistant_message_function_call = delta.delta.function_call
+
+                    # extract tool calls from response
+                    if assistant_message_tool_calls:
+                        tool_calls = self._extract_response_tool_calls(assistant_message_tool_calls)
+                        increase_tool_call(tool_calls)
+
+                    if delta_content is None or delta_content == "":
+                        continue
+
+                    # Add markdown quote markers for reasoning content
+                    if is_reasoning:
+                        if not is_reasoning_started:
+                            delta_content = "> 💭 " + delta_content
+                            is_reasoning_started = True
+                        elif "\n\n" in delta_content:
+                            delta_content = delta_content.replace("\n\n", "\n> ")
+                        elif "\n" in delta_content:
+                            delta_content = delta_content.replace("\n", "\n> ")
+                    elif is_reasoning_started:
+                        # If we were in reasoning mode but now getting regular content,
+                        # add \n\n to close the reasoning block
+                        delta_content = "\n\n" + delta_content
+                        is_reasoning_started = False
+
+                    # transform assistant message to prompt message
+                    assistant_prompt_message = AssistantPromptMessage(
+                        content=delta_content,
+                    )
+
+                    # reset tool calls
+                    tool_calls = []
+                    full_assistant_content += delta_content
+                elif "text" in choice:
+                    choice_text = choice.get("text", "")
+                    if choice_text == "":
+                        continue
+
+                    # transform assistant message to prompt message
+                    assistant_prompt_message = AssistantPromptMessage(content=choice_text)
+                    full_assistant_content += choice_text
+                else:
+                    continue
+
+                yield LLMResultChunk(
+                    id=message_id,
+                    model=model,
+                    prompt_messages=prompt_messages,
+                    delta=LLMResultChunkDelta(
+                        index=chunk_index,
+                        message=assistant_prompt_message,
+                    ),
+                )
+
+            chunk_index += 1
+
+        if tools_calls:
+            yield LLMResultChunk(
+                id=message_id,
+                model=model,
+                prompt_messages=prompt_messages,
+                delta=LLMResultChunkDelta(
+                    index=chunk_index,
+                    message=AssistantPromptMessage(tool_calls=tools_calls, content=""),
+                ),
+            )
+
+        yield create_final_llm_result_chunk(
+            id=message_id,
+            index=chunk_index,
+            message=AssistantPromptMessage(content=""),
+            finish_reason=finish_reason,
+            usage=usage,
+        )

api/core/model_runtime/model_providers/novita/_assets/icon_l_en.svg

@@ -1,19 +1,11 @@
-<svg width="162" height="36" viewBox="0 0 162 36" fill="none" xmlns="http://www.w3.org/2000/svg">
-<path fill-rule="evenodd" clip-rule="evenodd" d="M2 0C0.895431 0 0 0.895432 0 2V29.1891C0 30.2937 0.895433 31.1891 2 31.1891H5.51171L16.0608 35.1377C16.7145 35.3824 17.4114 34.8991 17.4114 34.2012V11.3669C17.4114 10.533 16.894 9.78665 16.1131 9.49405L5.51171 5.52152H25.58V31.1891H29.0917C30.1963 31.1891 31.0917 30.2937 31.0917 29.1891V2C31.0917 0.895431 30.1963 0 29.0917 0H2ZM14.6022 23.7351C15.0558 23.956 15.4239 23.6812 15.4239 23.1185C15.4239 22.5557 15.0558 21.9204 14.6022 21.6995C14.1486 21.4775 13.7804 21.7545 13.7804 22.3161C13.7804 22.8777 14.1486 23.513 14.6022 23.7351Z" fill="white"/>
-<path fill-rule="evenodd" clip-rule="evenodd" d="M2 0C0.895431 0 0 0.895432 0 2V29.1891C0 30.2937 0.895433 31.1891 2 31.1891H5.51171L16.0608 35.1377C16.7145 35.3824 17.4114 34.8991 17.4114 34.2012V11.3669C17.4114 10.533 16.894 9.78665 16.1131 9.49405L5.51171 5.52152H25.58V31.1891H29.0917C30.1963 31.1891 31.0917 30.2937 31.0917 29.1891V2C31.0917 0.895431 30.1963 0 29.0917 0H2ZM14.6022 23.7351C15.0558 23.956 15.4239 23.6812 15.4239 23.1185C15.4239 22.5557 15.0558 21.9204 14.6022 21.6995C14.1486 21.4775 13.7804 21.7545 13.7804 22.3161C13.7804 22.8777 14.1486 23.513 14.6022 23.7351Z" fill="url(#paint0_linear_1473_71)"/>
-<path d="M55.9397 27.8804H59.0566V19.0803C59.0566 14.9105 56.381 12.7172 52.8228 12.7172C51.0023 12.7172 49.3197 13.4483 48.2991 14.6668V12.9609H45.1546V27.8804H48.2991V19.5406C48.2991 16.8059 49.8162 15.3978 52.1332 15.3978C54.4226 15.3978 55.9397 16.8059 55.9397 19.5406V27.8804Z" fill="#11101A"/>
-<path fill-rule="evenodd" clip-rule="evenodd" d="M69.7881 12.7172C74.1187 12.7172 77.539 15.7228 77.539 20.4071C77.539 25.0915 74.0083 28.1241 69.6502 28.1241C65.3196 28.1241 62.0372 25.0915 62.0372 20.4071C62.0372 15.7228 65.4575 12.7172 69.7881 12.7172ZM69.7342 15.3979C67.362 15.3979 65.2381 17.0225 65.2381 20.4071C65.2381 23.7918 67.2793 25.4435 69.6514 25.4435C71.996 25.4435 74.313 23.7918 74.313 20.4071C74.313 17.0225 72.0788 15.3979 69.7342 15.3979Z" fill="#11101A"/>
-<path d="M78.861 12.9609L84.6259 27.8804H88.3772L94.1697 12.9609H90.8321L86.5291 25.1185L82.2261 12.9609H78.861Z" fill="#11101A"/>
-<path fill-rule="evenodd" clip-rule="evenodd" d="M100.13 9.00761C100.13 10.1178 99.2477 10.9842 98.1443 10.9842C97.0134 10.9842 96.1308 10.1178 96.1308 9.00761C96.1308 7.89745 97.0134 7.03098 98.1443 7.03098C99.2477 7.03098 100.13 7.89745 100.13 9.00761ZM99.6882 27.8804H96.5437V12.9609H99.6882V27.8804Z" fill="#11101A"/>
-<path d="M104.322 23.7376C104.322 26.7702 106.004 27.8804 108.708 27.8804H111.19V25.308H109.259C107.935 25.308 107.494 24.8477 107.494 23.7376V15.479H111.19V12.9609H107.494V9.25128H104.322V12.9609H102.529V15.479H104.322V23.7376Z" fill="#11101A"/>
-<path fill-rule="evenodd" clip-rule="evenodd" d="M120.154 28.1241C116.209 28.1241 113.037 24.9561 113.037 20.353C113.037 15.7498 116.209 12.7172 120.209 12.7172C122.774 12.7172 124.539 13.9086 125.477 15.1271V12.9609H128.649V27.8804H125.477V25.6601C124.512 26.9327 122.691 28.1241 120.154 28.1241ZM120.87 25.4435C123.242 25.4435 125.476 23.6293 125.476 20.4071C125.476 17.212 123.242 15.3979 120.87 15.3979C118.526 15.3979 116.264 17.1308 116.264 20.353C116.264 23.5752 118.526 25.4435 120.87 25.4435Z" fill="#11101A"/>
-<path d="M136.043 26.0933C136.043 24.9832 135.16 24.1167 134.057 24.1167C132.926 24.1167 132.043 24.9832 132.043 26.0933C132.043 27.2035 132.926 28.07 134.057 28.07C135.16 28.07 136.043 27.2035 136.043 26.0933Z" fill="#11101A"/>
-<path fill-rule="evenodd" clip-rule="evenodd" d="M145.502 28.1241C141.558 28.1241 138.386 24.9561 138.386 20.353C138.386 15.7498 141.558 12.7172 145.557 12.7172C148.123 12.7172 149.888 13.9086 150.826 15.1271V12.9609H153.998V27.8804H150.826V25.6601C149.86 26.9327 148.04 28.1241 145.502 28.1241ZM146.219 25.4435C148.591 25.4435 150.825 23.6293 150.825 20.4071C150.825 17.212 148.591 15.3979 146.219 15.3979C143.874 15.3979 141.612 17.1308 141.612 20.353C141.612 23.5752 143.874 25.4435 146.219 25.4435Z" fill="#11101A"/>
-<path fill-rule="evenodd" clip-rule="evenodd" d="M161.722 9.00761C161.722 10.1178 160.84 10.9842 159.736 10.9842C158.605 10.9842 157.723 10.1178 157.723 9.00761C157.723 7.89745 158.605 7.03098 159.736 7.03098C160.84 7.03098 161.722 7.89745 161.722 9.00761ZM161.28 27.8804H158.136V12.9609H161.28V27.8804Z" fill="#11101A"/>
+<svg width="88" height="24" viewBox="0 0 88 24" fill="none" xmlns="http://www.w3.org/2000/svg">
+<g clip-path="url(#clip0_1923_1287)">
+<path d="M24 18.8323V18.8326H14.3246L9.16716 13.6751V18.8326H0V18.8314L9.16716 9.66422V4H9.16774L24 18.8323Z" fill="black"/>
+</g>
+<path fill-rule="evenodd" clip-rule="evenodd" d="M73.2505 16.8061H76.5869V18.9145H73.9391C72.0857 18.9145 70.9202 17.8952 70.9202 15.9977V10.3921H69.0316V8.26609H70.9202L71.4677 5.47209H73.2329V8.26609H76.5869V10.3921H73.2505V16.8061ZM33.8133 4.85699L38.6679 15.681H38.809V4.85699H41.3333V18.9145H37.52L32.6654 8.09046H32.5243V18.9145H30V4.85699H33.8133ZM47.812 19.1254C44.7225 19.1254 42.7457 16.9641 42.7457 13.6079C42.7457 10.2517 44.6873 8.05518 47.812 8.05518C50.9367 8.05518 52.8429 10.1635 52.8429 13.6079C52.8429 17.0523 50.9014 19.1254 47.812 19.1254ZM47.812 17.017C49.1891 17.017 50.3363 16.5423 50.3715 15.1894V12.0265C50.3715 10.6383 49.2068 10.1635 47.812 10.1635C46.4172 10.1635 45.2171 10.6383 45.2171 12.0265V15.1894C45.2524 16.5599 46.4348 17.017 47.812 17.017ZM55.5444 8.24846L58.2979 16.6826H58.439L61.1926 8.24846H63.7346L59.9389 18.8968H56.7966L53.0186 8.24846H55.5429H55.5444ZM65.0419 8.26609H67.3722V18.9145H65.0419V8.26609ZM64.9001 4.85699H67.5126V6.86027H64.9001V4.85699ZM82.3064 19.143C79.4639 19.143 77.6458 16.9817 77.6458 13.6079C77.6458 10.2341 79.4286 8.07282 82.3064 8.07282C83.6483 8.07282 84.7425 8.59973 85.3958 9.58373H85.5369L85.9962 8.26609H87.7614V18.9145H85.9962L85.5369 17.6314H85.3958C84.6896 18.5625 83.5072 19.1423 82.3064 19.1423V19.143ZM82.7826 17.017C84.1774 17.017 85.3951 16.5776 85.4304 15.1894V12.0265C85.4304 10.603 84.159 10.1988 82.7297 10.1988C81.3004 10.1988 80.1172 10.6383 80.1172 12.0265V15.1894C80.1525 16.5952 81.3709 17.017 82.7826 17.017Z" fill="black"/>
 <defs>
-<linearGradient id="paint0_linear_1473_71" x1="31" y1="-2" x2="0.975591" y2="14.2625" gradientUnits="userSpaceOnUse">
-<stop stop-color="#2622FF"/>
-<stop offset="1" stop-color="#A717FF"/>
-</linearGradient>
+<clipPath id="clip0_1923_1287">
+<rect width="24" height="14.8326" fill="white" transform="translate(0 4)"/>
+</clipPath>
 </defs>
 </svg>

api/core/model_runtime/model_providers/novita/_assets/icon_s_en.svg

@@ -1,10 +1,3 @@
-<svg width="32" height="36" viewBox="0 0 32 36" fill="none" xmlns="http://www.w3.org/2000/svg">
-<path fill-rule="evenodd" clip-rule="evenodd" d="M2 0C0.895431 0 0 0.895432 0 2V29.1891C0 30.2937 0.895433 31.1891 2 31.1891H5.51171L16.0608 35.1377C16.7145 35.3824 17.4114 34.8991 17.4114 34.2012V11.3669C17.4114 10.533 16.894 9.78665 16.1131 9.49405L5.51171 5.52152H25.58V31.1891H29.0917C30.1963 31.1891 31.0917 30.2937 31.0917 29.1891V2C31.0917 0.895431 30.1963 0 29.0917 0H2ZM14.6022 23.7351C15.0558 23.956 15.4239 23.6812 15.4239 23.1185C15.4239 22.5557 15.0558 21.9204 14.6022 21.6995C14.1486 21.4775 13.7804 21.7545 13.7804 22.3161C13.7804 22.8777 14.1486 23.513 14.6022 23.7351Z" fill="white"/>
-<path fill-rule="evenodd" clip-rule="evenodd" d="M2 0C0.895431 0 0 0.895432 0 2V29.1891C0 30.2937 0.895433 31.1891 2 31.1891H5.51171L16.0608 35.1377C16.7145 35.3824 17.4114 34.8991 17.4114 34.2012V11.3669C17.4114 10.533 16.894 9.78665 16.1131 9.49405L5.51171 5.52152H25.58V31.1891H29.0917C30.1963 31.1891 31.0917 30.2937 31.0917 29.1891V2C31.0917 0.895431 30.1963 0 29.0917 0H2ZM14.6022 23.7351C15.0558 23.956 15.4239 23.6812 15.4239 23.1185C15.4239 22.5557 15.0558 21.9204 14.6022 21.6995C14.1486 21.4775 13.7804 21.7545 13.7804 22.3161C13.7804 22.8777 14.1486 23.513 14.6022 23.7351Z" fill="url(#paint0_linear_1473_97)"/>
-<defs>
-<linearGradient id="paint0_linear_1473_97" x1="31" y1="-2" x2="0.975591" y2="14.2625" gradientUnits="userSpaceOnUse">
-<stop stop-color="#2622FF"/>
-<stop offset="1" stop-color="#A717FF"/>
-</linearGradient>
-</defs>
+<svg width="24" height="15" viewBox="0 0 24 15" fill="none" xmlns="http://www.w3.org/2000/svg">
+<path d="M24 14.8323V14.8326H14.3246L9.16716 9.67507V14.8326H0V14.8314L9.16716 5.66422V0H9.16774L24 14.8323Z" fill="black"/>
 </svg>

api/core/model_runtime/model_providers/novita/llm/L3-8B-Stheno-v3.2.yaml

@@ -0,0 +1,41 @@
+model: Sao10K/L3-8B-Stheno-v3.2
+label:
+  zh_Hans: L3 8B Stheno V3.2
+  en_US: L3 8B Stheno V3.2
+model_type: llm
+features:
+  - agent-thought
+model_properties:
+  mode: chat
+  context_size: 8192
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+    min: 0
+    max: 2
+    default: 1
+  - name: top_p
+    use_template: top_p
+    min: 0
+    max: 1
+    default: 1
+  - name: max_tokens
+    use_template: max_tokens
+    min: 1
+    max: 2048
+    default: 512
+  - name: frequency_penalty
+    use_template: frequency_penalty
+    min: -2
+    max: 2
+    default: 0
+  - name: presence_penalty
+    use_template: presence_penalty
+    min: -2
+    max: 2
+    default: 0
+pricing:
+  input: '0.0005'
+  output: '0.0005'
+  unit: '0.0001'
+  currency: USD

api/core/model_runtime/model_providers/novita/llm/_position.yaml

@@ -0,0 +1,41 @@
+# Deepseek Models
+- deepseek/deepseek-r1
+- deepseek/deepseek_v3
+
+# LLaMA Models
+- meta-llama/llama-3.3-70b-instruct
+- meta-llama/llama-3.2-11b-vision-instruct
+- meta-llama/llama-3.2-3b-instruct
+- meta-llama/llama-3.2-1b-instruct
+- meta-llama/llama-3.1-70b-instruct
+- meta-llama/llama-3.1-8b-instruct
+- meta-llama/llama-3.1-8b-instruct-max
+- meta-llama/llama-3.1-8b-instruct-bf16
+- meta-llama/llama-3-70b-instruct
+- meta-llama/llama-3-8b-instruct
+
+# Mistral Models
+- mistralai/mistral-nemo
+- mistralai/mistral-7b-instruct
+
+# Qwen Models
+- qwen/qwen-2.5-72b-instruct
+- qwen/qwen-2-72b-instruct
+- qwen/qwen-2-vl-72b-instruct
+- qwen/qwen-2-7b-instruct
+
+# Other Models
+- sao10k/L3-8B-Stheno-v3.2
+- sao10k/l3-70b-euryale-v2.1
+- sao10k/l31-70b-euryale-v2.2
+- sao10k/l3-8b-lunaris
+- jondurbin/airoboros-l2-70b
+- cognitivecomputations/dolphin-mixtral-8x22b
+- google/gemma-2-9b-it
+- nousresearch/hermes-2-pro-llama-3-8b
+- sophosympatheia/midnight-rose-70b
+- gryphe/mythomax-l2-13b
+- nousresearch/nous-hermes-llama2-13b
+- openchat/openchat-7b
+- teknium/openhermes-2.5-mistral-7b
+- microsoft/wizardlm-2-8x22b

api/core/model_runtime/model_providers/novita/llm/airoboros-l2-70b.yaml

@@ -1,7 +1,7 @@
 model: jondurbin/airoboros-l2-70b
 label:
-  zh_Hans: jondurbin/airoboros-l2-70b
-  en_US: jondurbin/airoboros-l2-70b
+  zh_Hans: Airoboros L2 70B
+  en_US: Airoboros L2 70B
 model_type: llm
 features:
   - agent-thought

api/core/model_runtime/model_providers/novita/llm/deepseek-r1.yaml

@@ -0,0 +1,41 @@
+model: deepseek/deepseek-r1
+label:
+  zh_Hans: DeepSeek R1
+  en_US: DeepSeek R1
+model_type: llm
+features:
+  - agent-thought
+model_properties:
+  mode: chat
+  context_size: 64000
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+    min: 0
+    max: 2
+    default: 1
+  - name: top_p
+    use_template: top_p
+    min: 0
+    max: 1
+    default: 1
+  - name: max_tokens
+    use_template: max_tokens
+    min: 1
+    max: 2048
+    default: 512
+  - name: frequency_penalty
+    use_template: frequency_penalty
+    min: -2
+    max: 2
+    default: 0
+  - name: presence_penalty
+    use_template: presence_penalty
+    min: -2
+    max: 2
+    default: 0
+pricing:
+  input: '0.04'
+  output: '0.04'
+  unit: '0.0001'
+  currency: USD

api/core/model_runtime/model_providers/novita/llm/deepseek_v3.yaml

@@ -0,0 +1,41 @@
+model: deepseek/deepseek_v3
+label:
+  zh_Hans: DeepSeek V3
+  en_US: DeepSeek V3
+model_type: llm
+features:
+  - agent-thought
+model_properties:
+  mode: chat
+  context_size: 64000
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+    min: 0
+    max: 2
+    default: 1
+  - name: top_p
+    use_template: top_p
+    min: 0
+    max: 1
+    default: 1
+  - name: max_tokens
+    use_template: max_tokens
+    min: 1
+    max: 2048
+    default: 512
+  - name: frequency_penalty
+    use_template: frequency_penalty
+    min: -2
+    max: 2
+    default: 0
+  - name: presence_penalty
+    use_template: presence_penalty
+    min: -2
+    max: 2
+    default: 0
+pricing:
+  input: '0.0089'
+  output: '0.0089'
+  unit: '0.0001'
+  currency: USD

api/core/model_runtime/model_providers/novita/llm/dolphin-mixtral-8x22b.yaml

@@ -1,7 +1,7 @@
 model: cognitivecomputations/dolphin-mixtral-8x22b
 label:
-  zh_Hans: cognitivecomputations/dolphin-mixtral-8x22b
-  en_US: cognitivecomputations/dolphin-mixtral-8x22b
+  zh_Hans: Dolphin Mixtral 8x22B
+  en_US: Dolphin Mixtral 8x22B
 model_type: llm
 features:
   - agent-thought

api/core/model_runtime/model_providers/novita/llm/gemma-2-9b-it.yaml

@@ -1,7 +1,7 @@
 model: google/gemma-2-9b-it
 label:
-  zh_Hans: google/gemma-2-9b-it
-  en_US: google/gemma-2-9b-it
+  zh_Hans: Gemma 2 9B
+  en_US: Gemma 2 9B
 model_type: llm
 features:
   - agent-thought

api/core/model_runtime/model_providers/novita/llm/hermes-2-pro-llama-3-8b.yaml

@@ -1,7 +1,7 @@
 model: nousresearch/hermes-2-pro-llama-3-8b
 label:
-  zh_Hans: nousresearch/hermes-2-pro-llama-3-8b
-  en_US: nousresearch/hermes-2-pro-llama-3-8b
+  zh_Hans: Hermes 2 Pro Llama 3 8B
+  en_US: Hermes 2 Pro Llama 3 8B
 model_type: llm
 features:
   - agent-thought

api/core/model_runtime/model_providers/novita/llm/l3-70b-euryale-v2.1.yaml

@@ -1,7 +1,7 @@
 model: sao10k/l3-70b-euryale-v2.1
 label:
-  zh_Hans: sao10k/l3-70b-euryale-v2.1
-  en_US: sao10k/l3-70b-euryale-v2.1
+  zh_Hans: "L3 70B Euryale V2.1\t"
+  en_US: "L3 70B Euryale V2.1\t"
 model_type: llm
 features:
   - agent-thought

api/core/model_runtime/model_providers/novita/llm/l3-8b-lunaris.yaml

@@ -0,0 +1,41 @@
+model: sao10k/l3-8b-lunaris
+label:
+  zh_Hans: "Sao10k L3 8B Lunaris"
+  en_US: "Sao10k L3 8B Lunaris"
+model_type: llm
+features:
+  - agent-thought
+model_properties:
+  mode: chat
+  context_size: 8192
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+    min: 0
+    max: 2
+    default: 1
+  - name: top_p
+    use_template: top_p
+    min: 0
+    max: 1
+    default: 1
+  - name: max_tokens
+    use_template: max_tokens
+    min: 1
+    max: 2048
+    default: 512
+  - name: frequency_penalty
+    use_template: frequency_penalty
+    min: -2
+    max: 2
+    default: 0
+  - name: presence_penalty
+    use_template: presence_penalty
+    min: -2
+    max: 2
+    default: 0
+pricing:
+  input: '0.0005'
+  output: '0.0005'
+  unit: '0.0001'
+  currency: USD

api/core/model_runtime/model_providers/novita/llm/l31-70b-euryale-v2.2.yaml

@@ -0,0 +1,41 @@
+model: sao10k/l31-70b-euryale-v2.2
+label:
+  zh_Hans: L31 70B Euryale V2.2
+  en_US: L31 70B Euryale V2.2
+model_type: llm
+features:
+  - agent-thought
+model_properties:
+  mode: chat
+  context_size: 16000
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+    min: 0
+    max: 2
+    default: 1
+  - name: top_p
+    use_template: top_p
+    min: 0
+    max: 1
+    default: 1
+  - name: max_tokens
+    use_template: max_tokens
+    min: 1
+    max: 2048
+    default: 512
+  - name: frequency_penalty
+    use_template: frequency_penalty
+    min: -2
+    max: 2
+    default: 0
+  - name: presence_penalty
+    use_template: presence_penalty
+    min: -2
+    max: 2
+    default: 0
+pricing:
+  input: '0.0148'
+  output: '0.0148'
+  unit: '0.0001'
+  currency: USD

api/core/model_runtime/model_providers/novita/llm/llama-3-70b-instruct.yaml

@@ -1,7 +1,7 @@
 model: meta-llama/llama-3-70b-instruct
 label:
-  zh_Hans: meta-llama/llama-3-70b-instruct
-  en_US: meta-llama/llama-3-70b-instruct
+  zh_Hans: Llama3 70b Instruct
+  en_US: Llama3 70b Instruct
 model_type: llm
 features:
   - agent-thought

api/core/model_runtime/model_providers/novita/llm/llama-3-8b-instruct.yaml

@@ -1,7 +1,7 @@
 model: meta-llama/llama-3-8b-instruct
 label:
-  zh_Hans: meta-llama/llama-3-8b-instruct
-  en_US: meta-llama/llama-3-8b-instruct
+  zh_Hans: Llama 3 8B Instruct
+  en_US: Llama 3 8B Instruct
 model_type: llm
 features:
   - agent-thought
@@ -35,7 +35,7 @@ parameter_rules:
     max: 2
     default: 0
 pricing:
-  input: '0.00063'
-  output: '0.00063'
+  input: '0.0004'
+  output: '0.0004'
   unit: '0.0001'
   currency: USD

api/core/model_runtime/model_providers/novita/llm/llama-3.1-70b-instruct.yaml

@@ -1,13 +1,13 @@
 model: meta-llama/llama-3.1-70b-instruct
 label:
-  zh_Hans: meta-llama/llama-3.1-70b-instruct
-  en_US: meta-llama/llama-3.1-70b-instruct
+  zh_Hans: Llama 3.1 70B Instruct
+  en_US: Llama 3.1 70B Instruct
 model_type: llm
 features:
   - agent-thought
 model_properties:
   mode: chat
-  context_size: 8192
+  context_size: 32768
 parameter_rules:
   - name: temperature
     use_template: temperature
@@ -35,7 +35,7 @@ parameter_rules:
     max: 2
     default: 0
 pricing:
-  input: '0.0055'
-  output: '0.0076'
+  input: '0.0034'
+  output: '0.0039'
   unit: '0.0001'
   currency: USD

api/core/model_runtime/model_providers/novita/llm/llama-3.1-8b-instruct-bf16.yaml

@@ -0,0 +1,41 @@
+model: meta-llama/llama-3.1-8b-instruct-bf16
+label:
+  zh_Hans: Llama 3.1 8B Instruct BF16
+  en_US: Llama 3.1 8B Instruct BF16
+model_type: llm
+features:
+  - agent-thought
+model_properties:
+  mode: chat
+  context_size: 8192
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+    min: 0
+    max: 2
+    default: 1
+  - name: top_p
+    use_template: top_p
+    min: 0
+    max: 1
+    default: 1
+  - name: max_tokens
+    use_template: max_tokens
+    min: 1
+    max: 2048
+    default: 512
+  - name: frequency_penalty
+    use_template: frequency_penalty
+    min: -2
+    max: 2
+    default: 0
+  - name: presence_penalty
+    use_template: presence_penalty
+    min: -2
+    max: 2
+    default: 0
+pricing:
+  input: '0.0006'
+  output: '0.0006'
+  unit: '0.0001'
+  currency: USD

api/core/model_runtime/model_providers/novita/llm/llama-3.1-8b-instruct-max.yaml

@@ -0,0 +1,41 @@
+model: meta-llama/llama-3.1-8b-instruct-max
+label:
+  zh_Hans: "Llama3.1 8B Instruct Max\t"
+  en_US: "Llama3.1 8B Instruct Max\t"
+model_type: llm
+features:
+  - agent-thought
+model_properties:
+  mode: chat
+  context_size: 16384
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+    min: 0
+    max: 2
+    default: 1
+  - name: top_p
+    use_template: top_p
+    min: 0
+    max: 1
+    default: 1
+  - name: max_tokens
+    use_template: max_tokens
+    min: 1
+    max: 2048
+    default: 512
+  - name: frequency_penalty
+    use_template: frequency_penalty
+    min: -2
+    max: 2
+    default: 0
+  - name: presence_penalty
+    use_template: presence_penalty
+    min: -2
+    max: 2
+    default: 0
+pricing:
+  input: '0.0005'
+  output: '0.0005'
+  unit: '0.0001'
+  currency: USD

api/core/model_runtime/model_providers/novita/llm/llama-3.1-8b-instruct.yaml

@@ -1,13 +1,13 @@
 model: meta-llama/llama-3.1-8b-instruct
 label:
-  zh_Hans: meta-llama/llama-3.1-8b-instruct
-  en_US: meta-llama/llama-3.1-8b-instruct
+  zh_Hans: Llama 3.1 8B Instruct
+  en_US: Llama 3.1 8B Instruct
 model_type: llm
 features:
   - agent-thought
 model_properties:
   mode: chat
-  context_size: 8192
+  context_size: 16384
 parameter_rules:
   - name: temperature
     use_template: temperature
@@ -35,7 +35,7 @@ parameter_rules:
     max: 2
     default: 0
 pricing:
-  input: '0.001'
-  output: '0.001'
+  input: '0.0005'
+  output: '0.0005'
   unit: '0.0001'
   currency: USD

api/core/model_runtime/model_providers/novita/llm/llama-3.2-11b-vision-instruct.yaml

@@ -0,0 +1,41 @@
+model: meta-llama/llama-3.2-11b-vision-instruct
+label:
+  zh_Hans: "Llama 3.2 11B Vision Instruct\t"
+  en_US: "Llama 3.2 11B Vision Instruct\t"
+model_type: llm
+features:
+  - agent-thought
+model_properties:
+  mode: chat
+  context_size: 32768
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+    min: 0
+    max: 2
+    default: 1
+  - name: top_p
+    use_template: top_p
+    min: 0
+    max: 1
+    default: 1
+  - name: max_tokens
+    use_template: max_tokens
+    min: 1
+    max: 2048
+    default: 512
+  - name: frequency_penalty
+    use_template: frequency_penalty
+    min: -2
+    max: 2
+    default: 0
+  - name: presence_penalty
+    use_template: presence_penalty
+    min: -2
+    max: 2
+    default: 0
+pricing:
+  input: '0.0006'
+  output: '0.0006'
+  unit: '0.0001'
+  currency: USD

api/core/model_runtime/model_providers/novita/llm/llama-3.2-1b-instruct.yaml

@@ -0,0 +1,41 @@
+model: meta-llama/llama-3.2-1b-instruct
+label:
+  zh_Hans: "Llama 3.2 1B Instruct\t"
+  en_US: "Llama 3.2 1B Instruct\t"
+model_type: llm
+features:
+  - agent-thought
+model_properties:
+  mode: chat
+  context_size: 131000
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+    min: 0
+    max: 2
+    default: 1
+  - name: top_p
+    use_template: top_p
+    min: 0
+    max: 1
+    default: 1
+  - name: max_tokens
+    use_template: max_tokens
+    min: 1
+    max: 2048
+    default: 512
+  - name: frequency_penalty
+    use_template: frequency_penalty
+    min: -2
+    max: 2
+    default: 0
+  - name: presence_penalty
+    use_template: presence_penalty
+    min: -2
+    max: 2
+    default: 0
+pricing:
+  input: '0.0002'
+  output: '0.0002'
+  unit: '0.0001'
+  currency: USD

api/core/model_runtime/model_providers/novita/llm/llama-3.2-3b-instruct.yaml

@@ -1,7 +1,7 @@
-model: Nous-Hermes-2-Mixtral-8x7B-DPO
+model: meta-llama/llama-3.2-3b-instruct
 label:
-  zh_Hans: Nous-Hermes-2-Mixtral-8x7B-DPO
-  en_US: Nous-Hermes-2-Mixtral-8x7B-DPO
+  zh_Hans: Llama 3.2 3B Instruct
+  en_US: Llama 3.2 3B Instruct
 model_type: llm
 features:
   - agent-thought
@@ -35,7 +35,7 @@ parameter_rules:
     max: 2
     default: 0
 pricing:
-  input: '0.0027'
-  output: '0.0027'
+  input: '0.0003'
+  output: '0.0005'
   unit: '0.0001'
   currency: USD

api/core/model_runtime/model_providers/novita/llm/llama-3.3-70b-instruct.yaml

@@ -0,0 +1,41 @@
+model: meta-llama/llama-3.3-70b-instruct
+label:
+  zh_Hans: Llama 3.3 70B Instruct
+  en_US: Llama 3.3 70B Instruct
+model_type: llm
+features:
+  - agent-thought
+model_properties:
+  mode: chat
+  context_size: 131072
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+    min: 0
+    max: 2
+    default: 1
+  - name: top_p
+    use_template: top_p
+    min: 0
+    max: 1
+    default: 1
+  - name: max_tokens
+    use_template: max_tokens
+    min: 1
+    max: 2048
+    default: 512
+  - name: frequency_penalty
+    use_template: frequency_penalty
+    min: -2
+    max: 2
+    default: 0
+  - name: presence_penalty
+    use_template: presence_penalty
+    min: -2
+    max: 2
+    default: 0
+pricing:
+  input: '0.0039'
+  output: '0.0039'
+  unit: '0.0001'
+  currency: USD

api/core/model_runtime/model_providers/novita/llm/midnight-rose-70b.yaml

@@ -1,7 +1,7 @@
 model: sophosympatheia/midnight-rose-70b
 label:
-  zh_Hans: sophosympatheia/midnight-rose-70b
-  en_US: sophosympatheia/midnight-rose-70b
+  zh_Hans: Midnight Rose 70B
+  en_US: Midnight Rose 70B
 model_type: llm
 features:
   - agent-thought

api/core/model_runtime/model_providers/novita/llm/mistral-7b-instruct.yaml

@@ -1,7 +1,7 @@
 model: mistralai/mistral-7b-instruct
 label:
-  zh_Hans: mistralai/mistral-7b-instruct
-  en_US: mistralai/mistral-7b-instruct
+  zh_Hans: Mistral 7B Instruct
+  en_US: Mistral 7B Instruct
 model_type: llm
 features:
   - agent-thought

api/core/model_runtime/model_providers/novita/llm/mistral-nemo.yaml

@@ -0,0 +1,41 @@
+model: mistralai/mistral-nemo
+label:
+  zh_Hans: Mistral Nemo
+  en_US: Mistral Nemo
+model_type: llm
+features:
+  - agent-thought
+model_properties:
+  mode: chat
+  context_size: 131072
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+    min: 0
+    max: 2
+    default: 1
+  - name: top_p
+    use_template: top_p
+    min: 0
+    max: 1
+    default: 1
+  - name: max_tokens
+    use_template: max_tokens
+    min: 1
+    max: 2048
+    default: 512
+  - name: frequency_penalty
+    use_template: frequency_penalty
+    min: -2
+    max: 2
+    default: 0
+  - name: presence_penalty
+    use_template: presence_penalty
+    min: -2
+    max: 2
+    default: 0
+pricing:
+  input: '0.0017'
+  output: '0.0017'
+  unit: '0.0001'
+  currency: USD

api/core/model_runtime/model_providers/novita/llm/mythomax-l2-13b.yaml

@@ -1,7 +1,7 @@
 model: gryphe/mythomax-l2-13b
 label:
-  zh_Hans: gryphe/mythomax-l2-13b
-  en_US: gryphe/mythomax-l2-13b
+  zh_Hans: Mythomax L2 13B
+  en_US: Mythomax L2 13B
 model_type: llm
 features:
   - agent-thought
@@ -35,7 +35,7 @@ parameter_rules:
     max: 2
     default: 0
 pricing:
-  input: '0.00119'
-  output: '0.00119'
+  input: '0.0009'
+  output: '0.0009'
   unit: '0.0001'
   currency: USD

api/core/model_runtime/model_providers/novita/llm/nous-hermes-llama2-13b.yaml

@@ -1,7 +1,7 @@
 model: nousresearch/nous-hermes-llama2-13b
 label:
-  zh_Hans: nousresearch/nous-hermes-llama2-13b
-  en_US: nousresearch/nous-hermes-llama2-13b
+  zh_Hans: Nous Hermes Llama2 13B
+  en_US: Nous Hermes Llama2 13B
 model_type: llm
 features:
   - agent-thought

api/core/model_runtime/model_providers/novita/llm/openchat-7b.yaml

@@ -1,7 +1,7 @@
-model: lzlv_70b
+model: openchat/openchat-7b
 label:
-  zh_Hans: lzlv_70b
-  en_US: lzlv_70b
+  zh_Hans: OpenChat 7B
+  en_US: OpenChat 7B
 model_type: llm
 features:
   - agent-thought
@@ -35,7 +35,7 @@ parameter_rules:
     max: 2
     default: 0
 pricing:
-  input: '0.0058'
-  output: '0.0078'
+  input: '0.0006'
+  output: '0.0006'
   unit: '0.0001'
   currency: USD

api/core/model_runtime/model_providers/novita/llm/openhermes-2.5-mistral-7b.yaml

@@ -1,7 +1,7 @@
 model: teknium/openhermes-2.5-mistral-7b
 label:
-  zh_Hans: teknium/openhermes-2.5-mistral-7b
-  en_US: teknium/openhermes-2.5-mistral-7b
+  zh_Hans: Openhermes2.5 Mistral 7B
+  en_US: Openhermes2.5 Mistral 7B
 model_type: llm
 features:
   - agent-thought

api/core/model_runtime/model_providers/novita/llm/qwen-2-72b-instruct.yaml

@@ -1,7 +1,7 @@
-model: meta-llama/llama-3.1-405b-instruct
+model: qwen/qwen-2-72b-instruct
 label:
-  zh_Hans: meta-llama/llama-3.1-405b-instruct
-  en_US: meta-llama/llama-3.1-405b-instruct
+  zh_Hans: Qwen2 72B Instruct
+  en_US: Qwen2 72B Instruct
 model_type: llm
 features:
   - agent-thought
@@ -35,7 +35,7 @@ parameter_rules:
     max: 2
     default: 0
 pricing:
-  input: '0.03'
-  output: '0.05'
+  input: '0.0034'
+  output: '0.0039'
   unit: '0.0001'
   currency: USD

api/core/model_runtime/model_providers/novita/llm/qwen-2-7b-instruct.yaml

@@ -0,0 +1,41 @@
+model: qwen/qwen-2-7b-instruct
+label:
+  zh_Hans: Qwen 2 7B Instruct
+  en_US: Qwen 2 7B Instruct
+model_type: llm
+features:
+  - agent-thought
+model_properties:
+  mode: chat
+  context_size: 32768
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+    min: 0
+    max: 2
+    default: 1
+  - name: top_p
+    use_template: top_p
+    min: 0
+    max: 1
+    default: 1
+  - name: max_tokens
+    use_template: max_tokens
+    min: 1
+    max: 2048
+    default: 512
+  - name: frequency_penalty
+    use_template: frequency_penalty
+    min: -2
+    max: 2
+    default: 0
+  - name: presence_penalty
+    use_template: presence_penalty
+    min: -2
+    max: 2
+    default: 0
+pricing:
+  input: '0.00054'
+  output: '0.00054'
+  unit: '0.0001'
+  currency: USD

api/core/model_runtime/model_providers/novita/llm/qwen-2-vl-72b-instruct.yaml

@@ -0,0 +1,41 @@
+model: qwen/qwen-2-vl-72b-instruct
+label:
+  zh_Hans: Qwen 2 VL 72B Instruct
+  en_US: Qwen 2 VL 72B Instruct
+model_type: llm
+features:
+  - agent-thought
+model_properties:
+  mode: chat
+  context_size: 32768
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+    min: 0
+    max: 2
+    default: 1
+  - name: top_p
+    use_template: top_p
+    min: 0
+    max: 1
+    default: 1
+  - name: max_tokens
+    use_template: max_tokens
+    min: 1
+    max: 2048
+    default: 512
+  - name: frequency_penalty
+    use_template: frequency_penalty
+    min: -2
+    max: 2
+    default: 0
+  - name: presence_penalty
+    use_template: presence_penalty
+    min: -2
+    max: 2
+    default: 0
+pricing:
+  input: '0.0045'
+  output: '0.0045'
+  unit: '0.0001'
+  currency: USD

api/core/model_runtime/model_providers/novita/llm/qwen-2.5-72b-instruct.yaml

@@ -0,0 +1,41 @@
+model: qwen/qwen-2.5-72b-instruct
+label:
+  zh_Hans: Qwen 2.5 72B Instruct
+  en_US: Qwen 2.5 72B Instruct
+model_type: llm
+features:
+  - agent-thought
+model_properties:
+  mode: chat
+  context_size: 32000
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+    min: 0
+    max: 2
+    default: 1
+  - name: top_p
+    use_template: top_p
+    min: 0
+    max: 1
+    default: 1
+  - name: max_tokens
+    use_template: max_tokens
+    min: 1
+    max: 2048
+    default: 512
+  - name: frequency_penalty
+    use_template: frequency_penalty
+    min: -2
+    max: 2
+    default: 0
+  - name: presence_penalty
+    use_template: presence_penalty
+    min: -2
+    max: 2
+    default: 0
+pricing:
+  input: '0.0038'
+  output: '0.004'
+  unit: '0.0001'
+  currency: USD

api/core/model_runtime/model_providers/novita/llm/wizardlm-2-8x22b.yaml

@@ -1,7 +1,7 @@
 model: microsoft/wizardlm-2-8x22b
 label:
-  zh_Hans: microsoft/wizardlm-2-8x22b
-  en_US: microsoft/wizardlm-2-8x22b
+  zh_Hans: Wizardlm 2 8x22B
+  en_US: Wizardlm 2 8x22B
 model_type: llm
 features:
   - agent-thought
@@ -35,7 +35,7 @@ parameter_rules:
     max: 2
     default: 0
 pricing:
-  input: '0.0064'
-  output: '0.0064'
+  input: '0.0062'
+  output: '0.0062'
   unit: '0.0001'
   currency: USD

api/core/model_runtime/model_providers/novita/novita.yaml

@@ -1,6 +1,6 @@
 provider: novita
 label:
-  en_US: novita.ai
+  en_US: Novita AI
 description:
   en_US: An LLM API that matches various application scenarios with high cost-effectiveness.
   zh_Hans: 适配多种海外应用场景的高性价比 LLM API
@@ -8,13 +8,13 @@ icon_small:
   en_US: icon_s_en.svg
 icon_large:
   en_US: icon_l_en.svg
-background: "#eadeff"
+background: "#c7fce2"
 help:
   title:
-    en_US: Get your API key from novita.ai
-    zh_Hans: 从 novita.ai 获取 API Key
+    en_US: Get your API key from Novita AI
+    zh_Hans: 从 Novita AI 获取 API Key
   url:
-    en_US: https://novita.ai/settings#key-management?utm_source=dify&utm_medium=ch&utm_campaign=api
+    en_US: https://novita.ai/settings/key-management?utm_source=dify&utm_medium=ch&utm_campaign=api
 supported_model_types:
   - llm
 configurate_methods:

api/core/model_runtime/model_providers/openai/llm/llm.py

@@ -341,9 +341,6 @@ class OpenAILargeLanguageModel(_CommonOpenAI, LargeLanguageModel):
         :param credentials: provider credentials
         :return:
         """
-        # get predefined models
-        predefined_models = self.predefined_models()
-        predefined_models_map = {model.model: model for model in predefined_models}
 
         # transform credentials to kwargs for model instance
         credentials_kwargs = self._to_credential_kwargs(credentials)
@@ -359,9 +356,10 @@ class OpenAILargeLanguageModel(_CommonOpenAI, LargeLanguageModel):
             base_model = model.id.split(":")[1]
 
             base_model_schema = None
-            for predefined_model_name, predefined_model in predefined_models_map.items():
-                if predefined_model_name in base_model:
+            for predefined_model in self.predefined_models():
+                if predefined_model.model in base_model:
                     base_model_schema = predefined_model
+                    break
 
             if not base_model_schema:
                 continue
@@ -1186,12 +1184,14 @@ class OpenAILargeLanguageModel(_CommonOpenAI, LargeLanguageModel):
             base_model = model.split(":")[1]
 
         # get model schema
-        models = self.predefined_models()
-        model_map = {model.model: model for model in models}
-        if base_model not in model_map:
-            raise ValueError(f"Base model {base_model} not found")
+        base_model_schema = None
+        for predefined_model in self.predefined_models():
+            if base_model == predefined_model.model:
+                base_model_schema = predefined_model
+                break
 
-        base_model_schema = model_map[base_model]
+        if not base_model_schema:
+            raise ValueError(f"Base model {base_model} not found")
 
         base_model_schema_features = base_model_schema.features or []
         base_model_schema_model_properties = base_model_schema.model_properties

api/core/model_runtime/model_providers/perfxcloud/text_embedding/text_embedding.py

@@ -1,29 +1,13 @@
-import json
-import time
-from decimal import Decimal
 from typing import Optional
-from urllib.parse import urljoin
-
-import numpy as np
-import requests
 
 from core.entities.embedding_type import EmbeddingInputType
-from core.model_runtime.entities.common_entities import I18nObject
-from core.model_runtime.entities.model_entities import (
-    AIModelEntity,
-    FetchFrom,
-    ModelPropertyKey,
-    ModelType,
-    PriceConfig,
-    PriceType,
+from core.model_runtime.entities.text_embedding_entities import TextEmbeddingResult
+from core.model_runtime.model_providers.openai_api_compatible.text_embedding.text_embedding import (
+    OAICompatEmbeddingModel,
 )
-from core.model_runtime.entities.text_embedding_entities import EmbeddingUsage, TextEmbeddingResult
-from core.model_runtime.errors.validate import CredentialsValidateFailedError
-from core.model_runtime.model_providers.__base.text_embedding_model import TextEmbeddingModel
-from core.model_runtime.model_providers.openai_api_compatible._common import _CommonOaiApiCompat
 
 
-class OAICompatEmbeddingModel(_CommonOaiApiCompat, TextEmbeddingModel):
+class PerfXCloudEmbeddingModel(OAICompatEmbeddingModel):
     """
     Model class for an OpenAI API-compatible text embedding model.
     """
@@ -47,86 +31,10 @@ class OAICompatEmbeddingModel(_CommonOaiApiCompat, TextEmbeddingModel):
         :return: embeddings result
         """
 
-        # Prepare headers and payload for the request
-        headers = {"Content-Type": "application/json"}
-
-        api_key = credentials.get("api_key")
-        if api_key:
-            headers["Authorization"] = f"Bearer {api_key}"
-        endpoint_url: Optional[str]
         if "endpoint_url" not in credentials or credentials["endpoint_url"] == "":
-            endpoint_url = "https://cloud.perfxlab.cn/v1/"
-        else:
-            endpoint_url = credentials.get("endpoint_url")
-            assert endpoint_url is not None, "endpoint_url is required in credentials"
-            if not endpoint_url.endswith("/"):
-                endpoint_url += "/"
+            credentials["endpoint_url"] = "https://cloud.perfxlab.cn/v1/"
 
-        assert isinstance(endpoint_url, str)
-        endpoint_url = urljoin(endpoint_url, "embeddings")
-
-        extra_model_kwargs = {}
-        if user:
-            extra_model_kwargs["user"] = user
-
-        extra_model_kwargs["encoding_format"] = "float"
-
-        # get model properties
-        context_size = self._get_context_size(model, credentials)
-        max_chunks = self._get_max_chunks(model, credentials)
-
-        inputs = []
-        indices = []
-        used_tokens = 0
-
-        for i, text in enumerate(texts):
-            # Here token count is only an approximation based on the GPT2 tokenizer
-            # TODO: Optimize for better token estimation and chunking
-            num_tokens = self._get_num_tokens_by_gpt2(text)
-
-            if num_tokens >= context_size:
-                cutoff = int(np.floor(len(text) * (context_size / num_tokens)))
-                # if num tokens is larger than context length, only use the start
-                inputs.append(text[0:cutoff])
-            else:
-                inputs.append(text)
-            indices += [i]
-
-        batched_embeddings = []
-        _iter = range(0, len(inputs), max_chunks)
-
-        for i in _iter:
-            # Prepare the payload for the request
-            payload = {"input": inputs[i : i + max_chunks], "model": model, **extra_model_kwargs}
-
-            # Make the request to the OpenAI API
-            response = requests.post(endpoint_url, headers=headers, data=json.dumps(payload), timeout=(10, 300))
-
-            response.raise_for_status()  # Raise an exception for HTTP errors
-            response_data = response.json()
-
-            # Extract embeddings and used tokens from the response
-            embeddings_batch = [data["embedding"] for data in response_data["data"]]
-            embedding_used_tokens = response_data["usage"]["total_tokens"]
-
-            used_tokens += embedding_used_tokens
-            batched_embeddings += embeddings_batch
-
-        # calc usage
-        usage = self._calc_response_usage(model=model, credentials=credentials, tokens=used_tokens)
-
-        return TextEmbeddingResult(embeddings=batched_embeddings, usage=usage, model=model)
-
-    def get_num_tokens(self, model: str, credentials: dict, texts: list[str]) -> int:
-        """
-        Approximate number of tokens for given messages using GPT2 tokenizer
-
-        :param model: model name
-        :param credentials: model credentials
-        :param texts: texts to embed
-        :return:
-        """
-        return sum(self._get_num_tokens_by_gpt2(text) for text in texts)
+        return OAICompatEmbeddingModel._invoke(self, model, credentials, texts, user, input_type)
 
     def validate_credentials(self, model: str, credentials: dict) -> None:
         """
@@ -136,93 +44,7 @@ class OAICompatEmbeddingModel(_CommonOaiApiCompat, TextEmbeddingModel):
         :param credentials: model credentials
         :return:
         """
-        try:
-            headers = {"Content-Type": "application/json"}
+        if "endpoint_url" not in credentials or credentials["endpoint_url"] == "":
+            credentials["endpoint_url"] = "https://cloud.perfxlab.cn/v1/"
 
-            api_key = credentials.get("api_key")
-
-            if api_key:
-                headers["Authorization"] = f"Bearer {api_key}"
-
-            endpoint_url: Optional[str]
-            if "endpoint_url" not in credentials or credentials["endpoint_url"] == "":
-                endpoint_url = "https://cloud.perfxlab.cn/v1/"
-            else:
-                endpoint_url = credentials.get("endpoint_url")
-                assert endpoint_url is not None, "endpoint_url is required in credentials"
-                if not endpoint_url.endswith("/"):
-                    endpoint_url += "/"
-
-            assert isinstance(endpoint_url, str)
-            endpoint_url = urljoin(endpoint_url, "embeddings")
-
-            payload = {"input": "ping", "model": model}
-
-            response = requests.post(url=endpoint_url, headers=headers, data=json.dumps(payload), timeout=(10, 300))
-
-            if response.status_code != 200:
-                raise CredentialsValidateFailedError(
-                    f"Credentials validation failed with status code {response.status_code}"
-                )
-
-            try:
-                json_result = response.json()
-            except json.JSONDecodeError as e:
-                raise CredentialsValidateFailedError("Credentials validation failed: JSON decode error")
-
-            if "model" not in json_result:
-                raise CredentialsValidateFailedError("Credentials validation failed: invalid response")
-        except CredentialsValidateFailedError:
-            raise
-        except Exception as ex:
-            raise CredentialsValidateFailedError(str(ex))
-
-    def get_customizable_model_schema(self, model: str, credentials: dict) -> AIModelEntity:
-        """
-        generate custom model entities from credentials
-        """
-        entity = AIModelEntity(
-            model=model,
-            label=I18nObject(en_US=model),
-            model_type=ModelType.TEXT_EMBEDDING,
-            fetch_from=FetchFrom.CUSTOMIZABLE_MODEL,
-            model_properties={
-                ModelPropertyKey.CONTEXT_SIZE: int(credentials.get("context_size", 512)),
-                ModelPropertyKey.MAX_CHUNKS: 1,
-            },
-            parameter_rules=[],
-            pricing=PriceConfig(
-                input=Decimal(credentials.get("input_price", 0)),
-                unit=Decimal(credentials.get("unit", 0)),
-                currency=credentials.get("currency", "USD"),
-            ),
-        )
-
-        return entity
-
-    def _calc_response_usage(self, model: str, credentials: dict, tokens: int) -> EmbeddingUsage:
-        """
-        Calculate response usage
-
-        :param model: model name
-        :param credentials: model credentials
-        :param tokens: input tokens
-        :return: usage
-        """
-        # get input price info
-        input_price_info = self.get_price(
-            model=model, credentials=credentials, price_type=PriceType.INPUT, tokens=tokens
-        )
-
-        # transform usage
-        usage = EmbeddingUsage(
-            tokens=tokens,
-            total_tokens=tokens,
-            unit_price=input_price_info.unit_price,
-            price_unit=input_price_info.unit,
-            total_price=input_price_info.total_amount,
-            currency=input_price_info.currency,
-            latency=time.perf_counter() - self.started_at,
-        )
-
-        return usage
+        OAICompatEmbeddingModel.validate_credentials(self, model, credentials)

api/core/model_runtime/model_providers/siliconflow/llm/llm.py

@@ -1,9 +1,16 @@
+import json
 from collections.abc import Generator
 from typing import Optional, Union
 
+import requests
+
 from core.model_runtime.entities.common_entities import I18nObject
-from core.model_runtime.entities.llm_entities import LLMMode, LLMResult
-from core.model_runtime.entities.message_entities import PromptMessage, PromptMessageTool
+from core.model_runtime.entities.llm_entities import LLMMode, LLMResult, LLMResultChunk, LLMResultChunkDelta
+from core.model_runtime.entities.message_entities import (
+    AssistantPromptMessage,
+    PromptMessage,
+    PromptMessageTool,
+)
 from core.model_runtime.entities.model_entities import (
     AIModelEntity,
     FetchFrom,
@@ -89,3 +96,208 @@ class SiliconflowLargeLanguageModel(OAIAPICompatLargeLanguageModel):
                 ),
             ],
         )
+
+    def _handle_generate_stream_response(
+        self, model: str, credentials: dict, response: requests.Response, prompt_messages: list[PromptMessage]
+    ) -> Generator:
+        """
+        Handle llm stream response
+
+        :param model: model name
+        :param credentials: model credentials
+        :param response: streamed response
+        :param prompt_messages: prompt messages
+        :return: llm response chunk generator
+        """
+        full_assistant_content = ""
+        chunk_index = 0
+        is_reasoning_started = False  # Add flag to track reasoning state
+
+        def create_final_llm_result_chunk(
+            id: Optional[str], index: int, message: AssistantPromptMessage, finish_reason: str, usage: dict
+        ) -> LLMResultChunk:
+            # calculate num tokens
+            prompt_tokens = usage and usage.get("prompt_tokens")
+            if prompt_tokens is None:
+                prompt_tokens = self._num_tokens_from_string(model, prompt_messages[0].content)
+            completion_tokens = usage and usage.get("completion_tokens")
+            if completion_tokens is None:
+                completion_tokens = self._num_tokens_from_string(model, full_assistant_content)
+
+            # transform usage
+            usage = self._calc_response_usage(model, credentials, prompt_tokens, completion_tokens)
+
+            return LLMResultChunk(
+                id=id,
+                model=model,
+                prompt_messages=prompt_messages,
+                delta=LLMResultChunkDelta(index=index, message=message, finish_reason=finish_reason, usage=usage),
+            )
+
+        # delimiter for stream response, need unicode_escape
+        import codecs
+
+        delimiter = credentials.get("stream_mode_delimiter", "\n\n")
+        delimiter = codecs.decode(delimiter, "unicode_escape")
+
+        tools_calls: list[AssistantPromptMessage.ToolCall] = []
+
+        def increase_tool_call(new_tool_calls: list[AssistantPromptMessage.ToolCall]):
+            def get_tool_call(tool_call_id: str):
+                if not tool_call_id:
+                    return tools_calls[-1]
+
+                tool_call = next((tool_call for tool_call in tools_calls if tool_call.id == tool_call_id), None)
+                if tool_call is None:
+                    tool_call = AssistantPromptMessage.ToolCall(
+                        id=tool_call_id,
+                        type="function",
+                        function=AssistantPromptMessage.ToolCall.ToolCallFunction(name="", arguments=""),
+                    )
+                    tools_calls.append(tool_call)
+
+                return tool_call
+
+            for new_tool_call in new_tool_calls:
+                # get tool call
+                tool_call = get_tool_call(new_tool_call.function.name)
+                # update tool call
+                if new_tool_call.id:
+                    tool_call.id = new_tool_call.id
+                if new_tool_call.type:
+                    tool_call.type = new_tool_call.type
+                if new_tool_call.function.name:
+                    tool_call.function.name = new_tool_call.function.name
+                if new_tool_call.function.arguments:
+                    tool_call.function.arguments += new_tool_call.function.arguments
+
+        finish_reason = None  # The default value of finish_reason is None
+        message_id, usage = None, None
+        for chunk in response.iter_lines(decode_unicode=True, delimiter=delimiter):
+            chunk = chunk.strip()
+            if chunk:
+                # ignore sse comments
+                if chunk.startswith(":"):
+                    continue
+                decoded_chunk = chunk.strip().removeprefix("data:").lstrip()
+                if decoded_chunk == "[DONE]":  # Some provider returns "data: [DONE]"
+                    continue
+
+                try:
+                    chunk_json: dict = json.loads(decoded_chunk)
+                # stream ended
+                except json.JSONDecodeError as e:
+                    yield create_final_llm_result_chunk(
+                        id=message_id,
+                        index=chunk_index + 1,
+                        message=AssistantPromptMessage(content=""),
+                        finish_reason="Non-JSON encountered.",
+                        usage=usage,
+                    )
+                    break
+                # handle the error here. for issue #11629
+                if chunk_json.get("error") and chunk_json.get("choices") is None:
+                    raise ValueError(chunk_json.get("error"))
+
+                if chunk_json:
+                    if u := chunk_json.get("usage"):
+                        usage = u
+                if not chunk_json or len(chunk_json["choices"]) == 0:
+                    continue
+
+                choice = chunk_json["choices"][0]
+                finish_reason = chunk_json["choices"][0].get("finish_reason")
+                message_id = chunk_json.get("id")
+                chunk_index += 1
+
+                if "delta" in choice:
+                    delta = choice["delta"]
+                    delta_content = delta.get("content")
+
+                    assistant_message_tool_calls = None
+
+                    if "tool_calls" in delta and credentials.get("function_calling_type", "no_call") == "tool_call":
+                        assistant_message_tool_calls = delta.get("tool_calls", None)
+                    elif (
+                        "function_call" in delta
+                        and credentials.get("function_calling_type", "no_call") == "function_call"
+                    ):
+                        assistant_message_tool_calls = [
+                            {"id": "tool_call_id", "type": "function", "function": delta.get("function_call", {})}
+                        ]
+
+                    # assistant_message_function_call = delta.delta.function_call
+
+                    # extract tool calls from response
+                    if assistant_message_tool_calls:
+                        tool_calls = self._extract_response_tool_calls(assistant_message_tool_calls)
+                        increase_tool_call(tool_calls)
+
+                    if delta_content is None or delta_content == "":
+                        continue
+
+                    # Check for think tags
+                    if "<think>" in delta_content:
+                        is_reasoning_started = True
+                        # Remove <think> tag and add markdown quote
+                        delta_content = "> 💭 " + delta_content.replace("<think>", "")
+                    elif "</think>" in delta_content:
+                        # Remove </think> tag and add newlines to end quote block
+                        delta_content = delta_content.replace("</think>", "") + "\n\n"
+                        is_reasoning_started = False
+                    elif is_reasoning_started:
+                        # Add quote markers for content within thinking block
+                        if "\n\n" in delta_content:
+                            delta_content = delta_content.replace("\n\n", "\n> ")
+                        elif "\n" in delta_content:
+                            delta_content = delta_content.replace("\n", "\n> ")
+
+                    # transform assistant message to prompt message
+                    assistant_prompt_message = AssistantPromptMessage(
+                        content=delta_content,
+                    )
+
+                    # reset tool calls
+                    tool_calls = []
+                    full_assistant_content += delta_content
+                elif "text" in choice:
+                    choice_text = choice.get("text", "")
+                    if choice_text == "":
+                        continue
+
+                    # transform assistant message to prompt message
+                    assistant_prompt_message = AssistantPromptMessage(content=choice_text)
+                    full_assistant_content += choice_text
+                else:
+                    continue
+
+                yield LLMResultChunk(
+                    id=message_id,
+                    model=model,
+                    prompt_messages=prompt_messages,
+                    delta=LLMResultChunkDelta(
+                        index=chunk_index,
+                        message=assistant_prompt_message,
+                    ),
+                )
+
+            chunk_index += 1
+
+        if tools_calls:
+            yield LLMResultChunk(
+                id=message_id,
+                model=model,
+                prompt_messages=prompt_messages,
+                delta=LLMResultChunkDelta(
+                    index=chunk_index,
+                    message=AssistantPromptMessage(tool_calls=tools_calls, content=""),
+                ),
+            )
+
+        yield create_final_llm_result_chunk(
+            id=message_id,
+            index=chunk_index,
+            message=AssistantPromptMessage(content=""),
+            finish_reason=finish_reason,
+            usage=usage,
+        )

api/core/model_runtime/model_providers/tongyi/llm/_position.yaml

@@ -33,6 +33,8 @@
 - qwen2.5-3b-instruct
 - qwen2.5-1.5b-instruct
 - qwen2.5-0.5b-instruct
+- qwen2.5-14b-instruct-1m
+- qwen2.5-7b-instruct-1m
 - qwen2.5-coder-7b-instruct
 - qwen2-math-72b-instruct
 - qwen2-math-7b-instruct

api/core/model_runtime/model_providers/tongyi/llm/llm.py

@@ -219,8 +219,12 @@ class TongyiLargeLanguageModel(LargeLanguageModel):
         if response.status_code not in {200, HTTPStatus.OK}:
             raise ServiceUnavailableError(response.message)
         # transform assistant message to prompt message
+        resp_content = response.output.choices[0].message.content
+        # special for qwen-vl
+        if isinstance(resp_content, list):
+            resp_content = resp_content[0]["text"]
         assistant_prompt_message = AssistantPromptMessage(
-            content=response.output.choices[0].message.content,
+            content=resp_content,
         )
 
         # transform usage

api/core/model_runtime/model_providers/tongyi/llm/qwen2.5-14b-instruct-1m.yaml

@@ -0,0 +1,75 @@
+# for more details, please refer to https://help.aliyun.com/zh/model-studio/getting-started/models
+model: qwen2.5-14b-instruct-1m
+label:
+  en_US: qwen2.5-14b-instruct-1m
+model_type: llm
+features:
+  - agent-thought
+model_properties:
+  mode: chat
+  context_size: 1000000
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+    type: float
+    default: 0.3
+    min: 0.0
+    max: 2.0
+    help:
+      zh_Hans: 用于控制随机性和多样性的程度。具体来说，temperature值控制了生成文本时对每个候选词的概率分布进行平滑的程度。较高的temperature值会降低概率分布的峰值，使得更多的低概率词被选择，生成结果更加多样化；而较低的temperature值则会增强概率分布的峰值，使得高概率词更容易被选择，生成结果更加确定。
+      en_US: Used to control the degree of randomness and diversity. Specifically, the temperature value controls the degree to which the probability distribution of each candidate word is smoothed when generating text. A higher temperature value will reduce the peak value of the probability distribution, allowing more low-probability words to be selected, and the generated results will be more diverse; while a lower temperature value will enhance the peak value of the probability distribution, making it easier for high-probability words to be selected. , the generated results are more certain.
+  - name: max_tokens
+    use_template: max_tokens
+    type: int
+    default: 8192
+    min: 1
+    max: 8192
+    help:
+      zh_Hans: 用于指定模型在生成内容时token的最大数量，它定义了生成的上限，但不保证每次都会生成到这个数量。
+      en_US: It is used to specify the maximum number of tokens when the model generates content. It defines the upper limit of generation, but does not guarantee that this number will be generated every time.
+  - name: top_p
+    use_template: top_p
+    type: float
+    default: 0.8
+    min: 0.1
+    max: 0.9
+    help:
+      zh_Hans: 生成过程中核采样方法概率阈值，例如，取值为0.8时，仅保留概率加起来大于等于0.8的最可能token的最小集合作为候选集。取值范围为（0,1.0)，取值越大，生成的随机性越高；取值越低，生成的确定性越高。
+      en_US: The probability threshold of the kernel sampling method during the generation process. For example, when the value is 0.8, only the smallest set of the most likely tokens with a sum of probabilities greater than or equal to 0.8 is retained as the candidate set. The value range is (0,1.0). The larger the value, the higher the randomness generated; the lower the value, the higher the certainty generated.
+  - name: top_k
+    type: int
+    min: 0
+    max: 99
+    label:
+      zh_Hans: 取样数量
+      en_US: Top k
+    help:
+      zh_Hans: 生成时，采样候选集的大小。例如，取值为50时，仅将单次生成中得分最高的50个token组成随机采样的候选集。取值越大，生成的随机性越高；取值越小，生成的确定性越高。
+      en_US: The size of the sample candidate set when generated. For example, when the value is 50, only the 50 highest-scoring tokens in a single generation form a randomly sampled candidate set. The larger the value, the higher the randomness generated; the smaller the value, the higher the certainty generated.
+  - name: seed
+    required: false
+    type: int
+    default: 1234
+    label:
+      zh_Hans: 随机种子
+      en_US: Random seed
+    help:
+      zh_Hans: 生成时使用的随机数种子，用户控制模型生成内容的随机性。支持无符号64位整数，默认值为 1234。在使用seed时，模型将尽可能生成相同或相似的结果，但目前不保证每次生成的结果完全相同。
+      en_US: The random number seed used when generating, the user controls the randomness of the content generated by the model. Supports unsigned 64-bit integers, default value is 1234. When using seed, the model will try its best to generate the same or similar results, but there is currently no guarantee that the results will be exactly the same every time.
+  - name: repetition_penalty
+    required: false
+    type: float
+    default: 1.1
+    label:
+      zh_Hans: 重复惩罚
+      en_US: Repetition penalty
+    help:
+      zh_Hans: 用于控制模型生成时的重复度。提高repetition_penalty时可以降低模型生成的重复度。1.0表示不做惩罚。
+      en_US: Used to control the repeatability when generating models. Increasing repetition_penalty can reduce the duplication of model generation. 1.0 means no punishment.
+  - name: response_format
+    use_template: response_format
+pricing:
+  input: '0.001'
+  output: '0.003'
+  unit: '0.001'
+  currency: RMB

api/core/model_runtime/model_providers/tongyi/llm/qwen2.5-7b-instruct-1m.yaml

@@ -0,0 +1,75 @@
+# for more details, please refer to https://help.aliyun.com/zh/model-studio/getting-started/models
+model: qwen2.5-7b-instruct-1m
+label:
+  en_US: qwen2.5-7b-instruct-1m
+model_type: llm
+features:
+  - agent-thought
+model_properties:
+  mode: chat
+  context_size: 1000000
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+    type: float
+    default: 0.3
+    min: 0.0
+    max: 2.0
+    help:
+      zh_Hans: 用于控制随机性和多样性的程度。具体来说，temperature值控制了生成文本时对每个候选词的概率分布进行平滑的程度。较高的temperature值会降低概率分布的峰值，使得更多的低概率词被选择，生成结果更加多样化；而较低的temperature值则会增强概率分布的峰值，使得高概率词更容易被选择，生成结果更加确定。
+      en_US: Used to control the degree of randomness and diversity. Specifically, the temperature value controls the degree to which the probability distribution of each candidate word is smoothed when generating text. A higher temperature value will reduce the peak value of the probability distribution, allowing more low-probability words to be selected, and the generated results will be more diverse; while a lower temperature value will enhance the peak value of the probability distribution, making it easier for high-probability words to be selected. , the generated results are more certain.
+  - name: max_tokens
+    use_template: max_tokens
+    type: int
+    default: 8192
+    min: 1
+    max: 8192
+    help:
+      zh_Hans: 用于指定模型在生成内容时token的最大数量，它定义了生成的上限，但不保证每次都会生成到这个数量。
+      en_US: It is used to specify the maximum number of tokens when the model generates content. It defines the upper limit of generation, but does not guarantee that this number will be generated every time.
+  - name: top_p
+    use_template: top_p
+    type: float
+    default: 0.8
+    min: 0.1
+    max: 0.9
+    help:
+      zh_Hans: 生成过程中核采样方法概率阈值，例如，取值为0.8时，仅保留概率加起来大于等于0.8的最可能token的最小集合作为候选集。取值范围为（0,1.0)，取值越大，生成的随机性越高；取值越低，生成的确定性越高。
+      en_US: The probability threshold of the kernel sampling method during the generation process. For example, when the value is 0.8, only the smallest set of the most likely tokens with a sum of probabilities greater than or equal to 0.8 is retained as the candidate set. The value range is (0,1.0). The larger the value, the higher the randomness generated; the lower the value, the higher the certainty generated.
+  - name: top_k
+    type: int
+    min: 0
+    max: 99
+    label:
+      zh_Hans: 取样数量
+      en_US: Top k
+    help:
+      zh_Hans: 生成时，采样候选集的大小。例如，取值为50时，仅将单次生成中得分最高的50个token组成随机采样的候选集。取值越大，生成的随机性越高；取值越小，生成的确定性越高。
+      en_US: The size of the sample candidate set when generated. For example, when the value is 50, only the 50 highest-scoring tokens in a single generation form a randomly sampled candidate set. The larger the value, the higher the randomness generated; the smaller the value, the higher the certainty generated.
+  - name: seed
+    required: false
+    type: int
+    default: 1234
+    label:
+      zh_Hans: 随机种子
+      en_US: Random seed
+    help:
+      zh_Hans: 生成时使用的随机数种子，用户控制模型生成内容的随机性。支持无符号64位整数，默认值为 1234。在使用seed时，模型将尽可能生成相同或相似的结果，但目前不保证每次生成的结果完全相同。
+      en_US: The random number seed used when generating, the user controls the randomness of the content generated by the model. Supports unsigned 64-bit integers, default value is 1234. When using seed, the model will try its best to generate the same or similar results, but there is currently no guarantee that the results will be exactly the same every time.
+  - name: repetition_penalty
+    required: false
+    type: float
+    default: 1.1
+    label:
+      zh_Hans: 重复惩罚
+      en_US: Repetition penalty
+    help:
+      zh_Hans: 用于控制模型生成时的重复度。提高repetition_penalty时可以降低模型生成的重复度。1.0表示不做惩罚。
+      en_US: Used to control the repeatability when generating models. Increasing repetition_penalty can reduce the duplication of model generation. 1.0 means no punishment.
+  - name: response_format
+    use_template: response_format
+pricing:
+  input: '0.0005'
+  output: '0.001'
+  unit: '0.001'
+  currency: RMB

api/core/model_runtime/model_providers/zhipuai/llm/glm-4-air-0111.yaml

@@ -0,0 +1,66 @@
+model: glm-4-air-0111
+label:
+  en_US: glm-4-air-0111
+model_type: llm
+features:
+  - multi-tool-call
+  - agent-thought
+  - stream-tool-call
+model_properties:
+  mode: chat
+  context_size: 131072
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+    default: 0.95
+    min: 0.0
+    max: 1.0
+    help:
+      zh_Hans: 采样温度，控制输出的随机性，必须为正数取值范围是：(0.0,1.0]，不能等于 0,默认值为 0.95 值越大，会使输出更随机，更具创造性；值越小，输出会更加稳定或确定建议您根据应用场景调整 top_p 或 temperature 参数，但不要同时调整两个参数。
+      en_US: Sampling temperature, controls the randomness of the output, must be a positive number. The value range is (0.0,1.0], which cannot be equal to 0. The default value is 0.95. The larger the value, the more random and creative the output will be; the smaller the value, The output will be more stable or certain. It is recommended that you adjust the top_p or temperature parameters according to the application scenario, but do not adjust both parameters at the same time.
+  - name: top_p
+    use_template: top_p
+    default: 0.7
+    help:
+      zh_Hans: 用温度取样的另一种方法，称为核取样取值范围是：(0.0, 1.0) 开区间，不能等于 0 或 1，默认值为 0.7 模型考虑具有 top_p 概率质量tokens的结果例如：0.1 意味着模型解码器只考虑从前 10% 的概率的候选集中取 tokens 建议您根据应用场景调整 top_p 或 temperature 参数，但不要同时调整两个参数。
+      en_US: Another method of temperature sampling is called kernel sampling. The value range is (0.0, 1.0) open interval, which cannot be equal to 0 or 1. The default value is 0.7. The model considers the results with top_p probability mass tokens. For example 0.1 means The model decoder only considers tokens from the candidate set with the top 10% probability. It is recommended that you adjust the top_p or temperature parameters according to the application scenario, but do not adjust both parameters at the same time.
+  - name: do_sample
+    label:
+      zh_Hans: 采样策略
+      en_US: Sampling strategy
+    type: boolean
+    help:
+      zh_Hans: do_sample 为 true 时启用采样策略，do_sample 为 false 时采样策略 temperature、top_p 将不生效。默认值为 true。
+      en_US: When `do_sample` is set to true, the sampling strategy is enabled. When `do_sample` is set to false, the sampling strategies such as `temperature` and `top_p` will not take effect. The default value is true.
+    default: true
+  - name: max_tokens
+    use_template: max_tokens
+    default: 1024
+    min: 1
+    max: 4095
+  - name: web_search
+    type: boolean
+    label:
+      zh_Hans: 联网搜索
+      en_US: Web Search
+    default: false
+    help:
+      zh_Hans: 模型内置了互联网搜索服务，该参数控制模型在生成文本时是否参考使用互联网搜索结果。启用互联网搜索，模型会将搜索结果作为文本生成过程中的参考信息，但模型会基于其内部逻辑“自行判断”是否使用互联网搜索结果。
+      en_US: The model has a built-in Internet search service. This parameter controls whether the model refers to Internet search results when generating text. When Internet search is enabled, the model will use the search results as reference information in the text generation process, but the model will "judge" whether to use Internet search results based on its internal logic.
+  - name: response_format
+    label:
+      zh_Hans: 回复格式
+      en_US: Response Format
+    type: string
+    help:
+      zh_Hans: 指定模型必须输出的格式
+      en_US: specifying the format that the model must output
+    required: false
+    options:
+      - text
+      - json_object
+pricing:
+  input: '0.0005'
+  output: '0.0005'
+  unit: '0.001'
+  currency: RMB

api/core/rag/datasource/vdb/pgvector/pgvector.py

@@ -57,6 +57,11 @@ CREATE TABLE IF NOT EXISTS {table_name} (
 ) using heap;
 """
 
+SQL_CREATE_INDEX = """
+CREATE INDEX IF NOT EXISTS embedding_cosine_v1_idx ON {table_name} 
+USING hnsw (embedding vector_cosine_ops) WITH (m = 16, ef_construction = 64);
+"""
+
 
 class PGVector(BaseVector):
     def __init__(self, collection_name: str, config: PGVectorConfig):
@@ -205,7 +210,10 @@ class PGVector(BaseVector):
             with self._get_cursor() as cur:
                 cur.execute("CREATE EXTENSION IF NOT EXISTS vector")
                 cur.execute(SQL_CREATE_TABLE.format(table_name=self.table_name, dimension=dimension))
-                # TODO: create index https://github.com/pgvector/pgvector?tab=readme-ov-file#indexing
+                # PG hnsw index only support 2000 dimension or less
+                # ref: https://github.com/pgvector/pgvector?tab=readme-ov-file#indexing
+                if dimension <= 2000:
+                    cur.execute(SQL_CREATE_INDEX.format(table_name=self.table_name))
             redis_client.set(collection_exist_cache_key, 1, ex=3600)
 
 

api/core/rag/extractor/firecrawl/firecrawl_app.py

@@ -1,6 +1,6 @@
 import json
 import time
-from typing import cast
+from typing import Any, cast
 
 import requests
 
@@ -14,48 +14,47 @@ class FirecrawlApp:
         if self.api_key is None and self.base_url == "https://api.firecrawl.dev":
             raise ValueError("No API key provided")
 
-    def scrape_url(self, url, params=None) -> dict:
-        headers = {"Content-Type": "application/json", "Authorization": f"Bearer {self.api_key}"}
-        json_data = {"url": url}
+    def scrape_url(self, url, params=None) -> dict[str, Any]:
+        # Documentation: https://docs.firecrawl.dev/api-reference/endpoint/scrape
+        headers = self._prepare_headers()
+        json_data = {
+            "url": url,
+            "formats": ["markdown"],
+            "onlyMainContent": True,
+            "timeout": 30000,
+        }
         if params:
             json_data.update(params)
-        response = requests.post(f"{self.base_url}/v0/scrape", headers=headers, json=json_data)
+        response = self._post_request(f"{self.base_url}/v1/scrape", json_data, headers)
         if response.status_code == 200:
             response_data = response.json()
-            if response_data["success"] == True:
-                data = response_data["data"]
-                return {
-                    "title": data.get("metadata").get("title"),
-                    "description": data.get("metadata").get("description"),
-                    "source_url": data.get("metadata").get("sourceURL"),
-                    "markdown": data.get("markdown"),
-                }
-            else:
-                raise Exception(f"Failed to scrape URL. Error: {response_data['error']}")
-
-        elif response.status_code in {402, 409, 500}:
-            error_message = response.json().get("error", "Unknown error occurred")
-            raise Exception(f"Failed to scrape URL. Status code: {response.status_code}. Error: {error_message}")
+            data = response_data["data"]
+            return self._extract_common_fields(data)
+        elif response.status_code in {402, 409, 500, 429, 408}:
+            self._handle_error(response, "scrape URL")
+            return {}  # Avoid additional exception after handling error
         else:
             raise Exception(f"Failed to scrape URL. Status code: {response.status_code}")
 
     def crawl_url(self, url, params=None) -> str:
+        # Documentation: https://docs.firecrawl.dev/api-reference/endpoint/crawl-post
         headers = self._prepare_headers()
         json_data = {"url": url}
         if params:
             json_data.update(params)
-        response = self._post_request(f"{self.base_url}/v0/crawl", json_data, headers)
+        response = self._post_request(f"{self.base_url}/v1/crawl", json_data, headers)
         if response.status_code == 200:
-            job_id = response.json().get("jobId")
+            # There's also another two fields in the response: "success" (bool) and "url" (str)
+            job_id = response.json().get("id")
             return cast(str, job_id)
         else:
             self._handle_error(response, "start crawl job")
             # FIXME: unreachable code for mypy
             return ""  # unreachable
 
-    def check_crawl_status(self, job_id) -> dict:
+    def check_crawl_status(self, job_id) -> dict[str, Any]:
         headers = self._prepare_headers()
-        response = self._get_request(f"{self.base_url}/v0/crawl/status/{job_id}", headers)
+        response = self._get_request(f"{self.base_url}/v1/crawl/{job_id}", headers)
         if response.status_code == 200:
             crawl_status_response = response.json()
             if crawl_status_response.get("status") == "completed":
@@ -66,42 +65,48 @@ class FirecrawlApp:
                 url_data_list = []
                 for item in data:
                     if isinstance(item, dict) and "metadata" in item and "markdown" in item:
-                        url_data = {
-                            "title": item.get("metadata", {}).get("title"),
-                            "description": item.get("metadata", {}).get("description"),
-                            "source_url": item.get("metadata", {}).get("sourceURL"),
-                            "markdown": item.get("markdown"),
-                        }
+                        url_data = self._extract_common_fields(item)
                         url_data_list.append(url_data)
                 if url_data_list:
                     file_key = "website_files/" + job_id + ".txt"
-                    if storage.exists(file_key):
-                        storage.delete(file_key)
-                    storage.save(file_key, json.dumps(url_data_list).encode("utf-8"))
-                return {
-                    "status": "completed",
-                    "total": crawl_status_response.get("total"),
-                    "current": crawl_status_response.get("current"),
-                    "data": url_data_list,
-                }
-
+                    try:
+                        if storage.exists(file_key):
+                            storage.delete(file_key)
+                        storage.save(file_key, json.dumps(url_data_list).encode("utf-8"))
+                    except Exception as e:
+                        raise Exception(f"Error saving crawl data: {e}")
+                return self._format_crawl_status_response("completed", crawl_status_response, url_data_list)
             else:
-                return {
-                    "status": crawl_status_response.get("status"),
-                    "total": crawl_status_response.get("total"),
-                    "current": crawl_status_response.get("current"),
-                    "data": [],
-                }
-
+                return self._format_crawl_status_response(
+                    crawl_status_response.get("status"), crawl_status_response, []
+                )
         else:
             self._handle_error(response, "check crawl status")
             # FIXME: unreachable code for mypy
             return {}  # unreachable
 
-    def _prepare_headers(self):
+    def _format_crawl_status_response(
+        self, status: str, crawl_status_response: dict[str, Any], url_data_list: list[dict[str, Any]]
+    ) -> dict[str, Any]:
+        return {
+            "status": status,
+            "total": crawl_status_response.get("total"),
+            "current": crawl_status_response.get("completed"),
+            "data": url_data_list,
+        }
+
+    def _extract_common_fields(self, item: dict[str, Any]) -> dict[str, Any]:
+        return {
+            "title": item.get("metadata", {}).get("title"),
+            "description": item.get("metadata", {}).get("description"),
+            "source_url": item.get("metadata", {}).get("sourceURL"),
+            "markdown": item.get("markdown"),
+        }
+
+    def _prepare_headers(self) -> dict[str, Any]:
         return {"Content-Type": "application/json", "Authorization": f"Bearer {self.api_key}"}
 
-    def _post_request(self, url, data, headers, retries=3, backoff_factor=0.5):
+    def _post_request(self, url, data, headers, retries=3, backoff_factor=0.5) -> requests.Response:
         for attempt in range(retries):
             response = requests.post(url, headers=headers, json=data)
             if response.status_code == 502:
@@ -110,7 +115,7 @@ class FirecrawlApp:
                 return response
         return response
 
-    def _get_request(self, url, headers, retries=3, backoff_factor=0.5):
+    def _get_request(self, url, headers, retries=3, backoff_factor=0.5) -> requests.Response:
         for attempt in range(retries):
             response = requests.get(url, headers=headers)
             if response.status_code == 502:
@@ -119,6 +124,6 @@ class FirecrawlApp:
                 return response
         return response
 
-    def _handle_error(self, response, action):
+    def _handle_error(self, response, action) -> None:
         error_message = response.json().get("error", "Unknown error occurred")
         raise Exception(f"Failed to {action}. Status code: {response.status_code}. Error: {error_message}")

api/core/rag/extractor/firecrawl/firecrawl_web_extractor.py

@@ -13,9 +13,10 @@ class FirecrawlWebExtractor(BaseExtractor):
         api_key: The API key for Firecrawl.
         base_url: The base URL for the Firecrawl API. Defaults to 'https://api.firecrawl.dev'.
         mode: The mode of operation. Defaults to 'scrape'. Options are 'crawl', 'scrape' and 'crawl_return_urls'.
+        only_main_content: Only return the main content of the page excluding headers, navs, footers, etc.
     """
 
-    def __init__(self, url: str, job_id: str, tenant_id: str, mode: str = "crawl", only_main_content: bool = False):
+    def __init__(self, url: str, job_id: str, tenant_id: str, mode: str = "crawl", only_main_content: bool = True):
         """Initialize with url, api_key, base_url and mode."""
         self._url = url
         self.job_id = job_id

api/core/tools/provider/builtin/aws/tools/bedrock_config.py

@@ -0,0 +1,114 @@
+"""
+Configuration classes for AWS Bedrock retrieve and generate API
+"""
+
+from dataclasses import dataclass
+from typing import Any, Literal, Optional
+
+
+@dataclass
+class TextInferenceConfig:
+    """Text inference configuration"""
+
+    maxTokens: Optional[int] = None
+    stopSequences: Optional[list[str]] = None
+    temperature: Optional[float] = None
+    topP: Optional[float] = None
+
+
+@dataclass
+class PerformanceConfig:
+    """Performance configuration"""
+
+    latency: Literal["standard", "optimized"]
+
+
+@dataclass
+class PromptTemplate:
+    """Prompt template configuration"""
+
+    textPromptTemplate: str
+
+
+@dataclass
+class GuardrailConfig:
+    """Guardrail configuration"""
+
+    guardrailId: str
+    guardrailVersion: str
+
+
+@dataclass
+class GenerationConfig:
+    """Generation configuration"""
+
+    additionalModelRequestFields: Optional[dict[str, Any]] = None
+    guardrailConfiguration: Optional[GuardrailConfig] = None
+    inferenceConfig: Optional[dict[str, TextInferenceConfig]] = None
+    performanceConfig: Optional[PerformanceConfig] = None
+    promptTemplate: Optional[PromptTemplate] = None
+
+
+@dataclass
+class VectorSearchConfig:
+    """Vector search configuration"""
+
+    filter: Optional[dict[str, Any]] = None
+    numberOfResults: Optional[int] = None
+    overrideSearchType: Optional[Literal["HYBRID", "SEMANTIC"]] = None
+
+
+@dataclass
+class RetrievalConfig:
+    """Retrieval configuration"""
+
+    vectorSearchConfiguration: VectorSearchConfig
+
+
+@dataclass
+class OrchestrationConfig:
+    """Orchestration configuration"""
+
+    additionalModelRequestFields: Optional[dict[str, Any]] = None
+    inferenceConfig: Optional[dict[str, TextInferenceConfig]] = None
+    performanceConfig: Optional[PerformanceConfig] = None
+    promptTemplate: Optional[PromptTemplate] = None
+
+
+@dataclass
+class KnowledgeBaseConfig:
+    """Knowledge base configuration"""
+
+    generationConfiguration: GenerationConfig
+    knowledgeBaseId: str
+    modelArn: str
+    orchestrationConfiguration: Optional[OrchestrationConfig] = None
+    retrievalConfiguration: Optional[RetrievalConfig] = None
+
+
+@dataclass
+class SessionConfig:
+    """Session configuration"""
+
+    kmsKeyArn: Optional[str] = None
+    sessionId: Optional[str] = None
+
+
+@dataclass
+class RetrieveAndGenerateConfiguration:
+    """Retrieve and generate configuration
+    The use of knowledgeBaseConfiguration or externalSourcesConfiguration depends on the type value
+    """
+
+    type: str = "KNOWLEDGE_BASE"
+    knowledgeBaseConfiguration: Optional[KnowledgeBaseConfig] = None
+
+
+@dataclass
+class RetrieveAndGenerateConfig:
+    """Retrieve and generate main configuration"""
+
+    input: dict[str, str]
+    retrieveAndGenerateConfiguration: RetrieveAndGenerateConfiguration
+    sessionConfiguration: Optional[SessionConfig] = None
+    sessionId: Optional[str] = None

api/core/tools/provider/builtin/aws/tools/bedrock_retrieve_and_generate.py

@@ -0,0 +1,324 @@
+import json
+from typing import Any, Optional
+
+import boto3
+
+from core.tools.entities.tool_entities import ToolInvokeMessage
+from core.tools.tool.builtin_tool import BuiltinTool
+
+
+class BedrockRetrieveAndGenerateTool(BuiltinTool):
+    bedrock_client: Any = None
+
+    def _create_text_inference_config(
+        self,
+        max_tokens: Optional[int] = None,
+        stop_sequences: Optional[str] = None,
+        temperature: Optional[float] = None,
+        top_p: Optional[float] = None,
+    ) -> Optional[dict]:
+        """Create text inference configuration"""
+        if any([max_tokens, stop_sequences, temperature, top_p]):
+            config = {}
+            if max_tokens is not None:
+                config["maxTokens"] = max_tokens
+            if stop_sequences:
+                try:
+                    config["stopSequences"] = json.loads(stop_sequences)
+                except json.JSONDecodeError:
+                    config["stopSequences"] = []
+            if temperature is not None:
+                config["temperature"] = temperature
+            if top_p is not None:
+                config["topP"] = top_p
+            return config
+        return None
+
+    def _create_guardrail_config(
+        self,
+        guardrail_id: Optional[str] = None,
+        guardrail_version: Optional[str] = None,
+    ) -> Optional[dict]:
+        """Create guardrail configuration"""
+        if guardrail_id and guardrail_version:
+            return {"guardrailId": guardrail_id, "guardrailVersion": guardrail_version}
+        return None
+
+    def _create_generation_config(
+        self,
+        additional_model_fields: Optional[str] = None,
+        guardrail_config: Optional[dict] = None,
+        text_inference_config: Optional[dict] = None,
+        performance_mode: Optional[str] = None,
+        prompt_template: Optional[str] = None,
+    ) -> dict:
+        """Create generation configuration"""
+        config = {}
+
+        if additional_model_fields:
+            try:
+                config["additionalModelRequestFields"] = json.loads(additional_model_fields)
+            except json.JSONDecodeError:
+                pass
+
+        if guardrail_config:
+            config["guardrailConfiguration"] = guardrail_config
+
+        if text_inference_config:
+            config["inferenceConfig"] = {"textInferenceConfig": text_inference_config}
+
+        if performance_mode:
+            config["performanceConfig"] = {"latency": performance_mode}
+
+        if prompt_template:
+            config["promptTemplate"] = {"textPromptTemplate": prompt_template}
+
+        return config
+
+    def _create_orchestration_config(
+        self,
+        orchestration_additional_model_fields: Optional[str] = None,
+        orchestration_text_inference_config: Optional[dict] = None,
+        orchestration_performance_mode: Optional[str] = None,
+        orchestration_prompt_template: Optional[str] = None,
+    ) -> dict:
+        """Create orchestration configuration"""
+        config = {}
+
+        if orchestration_additional_model_fields:
+            try:
+                config["additionalModelRequestFields"] = json.loads(orchestration_additional_model_fields)
+            except json.JSONDecodeError:
+                pass
+
+        if orchestration_text_inference_config:
+            config["inferenceConfig"] = {"textInferenceConfig": orchestration_text_inference_config}
+
+        if orchestration_performance_mode:
+            config["performanceConfig"] = {"latency": orchestration_performance_mode}
+
+        if orchestration_prompt_template:
+            config["promptTemplate"] = {"textPromptTemplate": orchestration_prompt_template}
+
+        return config
+
+    def _create_vector_search_config(
+        self,
+        number_of_results: int = 5,
+        search_type: str = "SEMANTIC",
+        metadata_filter: Optional[dict] = None,
+    ) -> dict:
+        """Create vector search configuration"""
+        config = {
+            "numberOfResults": number_of_results,
+            "overrideSearchType": search_type,
+        }
+
+        # Only add filter if metadata_filter is not empty
+        if metadata_filter:
+            config["filter"] = metadata_filter
+
+        return config
+
+    def _bedrock_retrieve_and_generate(
+        self,
+        query: str,
+        knowledge_base_id: str,
+        model_arn: str,
+        # Generation Configuration
+        additional_model_fields: Optional[str] = None,
+        guardrail_id: Optional[str] = None,
+        guardrail_version: Optional[str] = None,
+        max_tokens: Optional[int] = None,
+        stop_sequences: Optional[str] = None,
+        temperature: Optional[float] = None,
+        top_p: Optional[float] = None,
+        performance_mode: str = "standard",
+        prompt_template: Optional[str] = None,
+        # Orchestration Configuration
+        orchestration_additional_model_fields: Optional[str] = None,
+        orchestration_max_tokens: Optional[int] = None,
+        orchestration_stop_sequences: Optional[str] = None,
+        orchestration_temperature: Optional[float] = None,
+        orchestration_top_p: Optional[float] = None,
+        orchestration_performance_mode: Optional[str] = None,
+        orchestration_prompt_template: Optional[str] = None,
+        # Retrieval Configuration
+        number_of_results: int = 5,
+        search_type: str = "SEMANTIC",
+        metadata_filter: Optional[dict] = None,
+        # Additional Configuration
+        session_id: Optional[str] = None,
+    ) -> dict[str, Any]:
+        try:
+            # Create text inference configurations
+            text_inference_config = self._create_text_inference_config(max_tokens, stop_sequences, temperature, top_p)
+            orchestration_text_inference_config = self._create_text_inference_config(
+                orchestration_max_tokens, orchestration_stop_sequences, orchestration_temperature, orchestration_top_p
+            )
+
+            # Create guardrail configuration
+            guardrail_config = self._create_guardrail_config(guardrail_id, guardrail_version)
+
+            # Create vector search configuration
+            vector_search_config = self._create_vector_search_config(number_of_results, search_type, metadata_filter)
+
+            # Create generation configuration
+            generation_config = self._create_generation_config(
+                additional_model_fields, guardrail_config, text_inference_config, performance_mode, prompt_template
+            )
+
+            # Create orchestration configuration
+            orchestration_config = self._create_orchestration_config(
+                orchestration_additional_model_fields,
+                orchestration_text_inference_config,
+                orchestration_performance_mode,
+                orchestration_prompt_template,
+            )
+
+            # Create knowledge base configuration
+            knowledge_base_config = {
+                "knowledgeBaseId": knowledge_base_id,
+                "modelArn": model_arn,
+                "generationConfiguration": generation_config,
+                "orchestrationConfiguration": orchestration_config,
+                "retrievalConfiguration": {"vectorSearchConfiguration": vector_search_config},
+            }
+
+            # Create request configuration
+            request_config = {
+                "input": {"text": query},
+                "retrieveAndGenerateConfiguration": {
+                    "type": "KNOWLEDGE_BASE",
+                    "knowledgeBaseConfiguration": knowledge_base_config,
+                },
+            }
+
+            # Add session configuration if provided
+            if session_id and len(session_id) >= 2:
+                request_config["sessionConfiguration"] = {"sessionId": session_id}
+                request_config["sessionId"] = session_id
+
+            # Send request
+            response = self.bedrock_client.retrieve_and_generate(**request_config)
+
+            # Process response
+            result = {"output": response.get("output", {}).get("text", ""), "citations": []}
+
+            # Process citations
+            for citation in response.get("citations", []):
+                citation_info = {
+                    "text": citation.get("generatedResponsePart", {}).get("textResponsePart", {}).get("text", ""),
+                    "references": [],
+                }
+
+                for ref in citation.get("retrievedReferences", []):
+                    reference = {
+                        "content": ref.get("content", {}).get("text", ""),
+                        "metadata": ref.get("metadata", {}),
+                        "location": None,
+                    }
+
+                    location = ref.get("location", {})
+                    if location.get("type") == "S3":
+                        reference["location"] = location.get("s3Location", {}).get("uri")
+
+                    citation_info["references"].append(reference)
+
+                result["citations"].append(citation_info)
+
+            return result
+
+        except Exception as e:
+            raise Exception(f"Error calling Bedrock service: {str(e)}")
+
+    def _invoke(
+        self,
+        user_id: str,
+        tool_parameters: dict[str, Any],
+    ) -> ToolInvokeMessage:
+        try:
+            # Initialize Bedrock client if not already initialized
+            if not self.bedrock_client:
+                aws_region = tool_parameters.get("aws_region")
+                aws_access_key_id = tool_parameters.get("aws_access_key_id")
+                aws_secret_access_key = tool_parameters.get("aws_secret_access_key")
+
+                client_kwargs = {
+                    "service_name": "bedrock-agent-runtime",
+                }
+                if aws_region:
+                    client_kwargs["region_name"] = aws_region
+                # Only add credentials if both access key and secret key are provided
+                if aws_access_key_id and aws_secret_access_key:
+                    client_kwargs.update(
+                        {"aws_access_key_id": aws_access_key_id, "aws_secret_access_key": aws_secret_access_key}
+                    )
+
+                try:
+                    self.bedrock_client = boto3.client(**client_kwargs)
+                except Exception as e:
+                    return self.create_text_message(f"Failed to initialize Bedrock client: {str(e)}")
+
+            # Parse metadata filter if provided
+            metadata_filter = None
+            if metadata_filter_str := tool_parameters.get("metadata_filter"):
+                try:
+                    parsed_filter = json.loads(metadata_filter_str)
+                    if parsed_filter:  # Only set if not empty
+                        metadata_filter = parsed_filter
+                except json.JSONDecodeError:
+                    return self.create_text_message("metadata_filter must be a valid JSON string")
+
+            try:
+                response = self._bedrock_retrieve_and_generate(
+                    query=tool_parameters["query"],
+                    knowledge_base_id=tool_parameters["knowledge_base_id"],
+                    model_arn=tool_parameters["model_arn"],
+                    # Generation Configuration
+                    additional_model_fields=tool_parameters.get("additional_model_fields"),
+                    guardrail_id=tool_parameters.get("guardrail_id"),
+                    guardrail_version=tool_parameters.get("guardrail_version"),
+                    max_tokens=tool_parameters.get("max_tokens"),
+                    stop_sequences=tool_parameters.get("stop_sequences"),
+                    temperature=tool_parameters.get("temperature"),
+                    top_p=tool_parameters.get("top_p"),
+                    performance_mode=tool_parameters.get("performance_mode", "standard"),
+                    prompt_template=tool_parameters.get("prompt_template"),
+                    # Orchestration Configuration
+                    orchestration_additional_model_fields=tool_parameters.get("orchestration_additional_model_fields"),
+                    orchestration_max_tokens=tool_parameters.get("orchestration_max_tokens"),
+                    orchestration_stop_sequences=tool_parameters.get("orchestration_stop_sequences"),
+                    orchestration_temperature=tool_parameters.get("orchestration_temperature"),
+                    orchestration_top_p=tool_parameters.get("orchestration_top_p"),
+                    orchestration_performance_mode=tool_parameters.get("orchestration_performance_mode"),
+                    orchestration_prompt_template=tool_parameters.get("orchestration_prompt_template"),
+                    # Retrieval Configuration
+                    number_of_results=tool_parameters.get("number_of_results", 5),
+                    search_type=tool_parameters.get("search_type", "SEMANTIC"),
+                    metadata_filter=metadata_filter,
+                    # Additional Configuration
+                    session_id=tool_parameters.get("session_id"),
+                )
+                return self.create_json_message(response)
+
+            except Exception as e:
+                return self.create_text_message(f"Tool invocation error: {str(e)}")
+
+        except Exception as e:
+            return self.create_text_message(f"Tool execution error: {str(e)}")
+
+    def validate_parameters(self, parameters: dict[str, Any]) -> None:
+        """Validate the parameters"""
+        required_params = ["query", "model_arn", "knowledge_base_id"]
+        for param in required_params:
+            if not parameters.get(param):
+                raise ValueError(f"{param} is required")
+
+        # Validate metadata filter if provided
+        if metadata_filter_str := parameters.get("metadata_filter"):
+            try:
+                if not isinstance(json.loads(metadata_filter_str), dict):
+                    raise ValueError("metadata_filter must be a valid JSON object")
+            except json.JSONDecodeError:
+                raise ValueError("metadata_filter must be a valid JSON string")

api/core/tools/provider/builtin/aws/tools/bedrock_retrieve_and_generate.yaml

@@ -0,0 +1,358 @@
+identity:
+  name: bedrock_retrieve_and_generate
+  author: AWS
+  label:
+    en_US: Bedrock Retrieve and Generate
+    zh_Hans: Bedrock检索和生成
+  icon: icon.svg
+
+description:
+  human:
+    en_US: A tool for retrieving and generating information using Amazon Bedrock Knowledge Base
+    zh_Hans: 使用Amazon Bedrock知识库进行信息检索和生成的工具
+  llm: A tool for retrieving and generating information using Amazon Bedrock Knowledge Base
+
+parameters:
+# Additional Configuration
+  - name: session_id
+    type: string
+    required: false
+    label:
+      en_US: Session ID
+      zh_Hans: 会话ID
+    human_description:
+      en_US: Optional session ID for continuous conversations
+      zh_Hans: 用于连续对话的可选会话ID
+    form: form
+
+  # AWS Configuration
+  - name: aws_region
+    type: string
+    required: false
+    label:
+      en_US: AWS Region
+      zh_Hans: AWS区域
+    human_description:
+      en_US: AWS region for the Bedrock service
+      zh_Hans: Bedrock服务的AWS区域
+    form: form
+
+  - name: aws_access_key_id
+    type: string
+    required: false
+    label:
+      en_US: AWS Access Key ID
+      zh_Hans: AWS访问密钥ID
+    human_description:
+      en_US: AWS access key ID for authentication (optional)
+      zh_Hans: 用于身份验证的AWS访问密钥ID（可选）
+    form: form
+
+  - name: aws_secret_access_key
+    type: string
+    required: false
+    label:
+      en_US: AWS Secret Access Key
+      zh_Hans: AWS秘密访问密钥
+    human_description:
+      en_US: AWS secret access key for authentication (optional)
+      zh_Hans: 用于身份验证的AWS秘密访问密钥（可选）
+    form: form
+
+  # Knowledge Base Configuration
+  - name: knowledge_base_id
+    type: string
+    required: true
+    label:
+      en_US: Knowledge Base ID
+      zh_Hans: 知识库ID
+    human_description:
+      en_US: ID of the Bedrock Knowledge Base
+      zh_Hans: Bedrock知识库的ID
+    form: form
+
+  - name: model_arn
+    type: string
+    required: true
+    label:
+      en_US: Model ARN
+      zh_Hans: 模型ARN
+    human_description:
+      en_US: The ARN of the model to use
+      zh_Hans: 要使用的模型ARN
+    form: form
+
+  # Retrieval Configuration
+  - name: query
+    type: string
+    required: true
+    label:
+      en_US: Query
+      zh_Hans: 查询
+    human_description:
+      en_US: The search query to retrieve information
+      zh_Hans: 用于检索信息的查询语句
+    form: llm
+
+  - name: number_of_results
+    type: number
+    required: false
+    label:
+      en_US: Number of Results
+      zh_Hans: 结果数量
+    human_description:
+      en_US: Number of results to retrieve (1-10)
+      zh_Hans: 要检索的结果数量（1-10）
+    default: 5
+    min: 1
+    max: 10
+    form: form
+
+  - name: search_type
+    type: select
+    required: false
+    label:
+      en_US: Search Type
+      zh_Hans: 搜索类型
+    human_description:
+      en_US: Type of search to perform
+      zh_Hans: 要执行的搜索类型
+    default: SEMANTIC
+    options:
+      - value: SEMANTIC
+        label:
+          en_US: Semantic Search
+          zh_Hans: 语义搜索
+      - value: HYBRID
+        label:
+          en_US: Hybrid Search
+          zh_Hans: 混合搜索
+    form: form
+
+  - name: metadata_filter
+    type: string
+    required: false
+    label:
+      en_US: Metadata Filter
+      zh_Hans: 元数据过滤器
+    human_description:
+      en_US: JSON formatted filter conditions for metadata, supporting operations like equals, greaterThan, lessThan, etc.
+      zh_Hans: 元数据的JSON格式过滤条件，支持等于、大于、小于等操作
+    default: "{}"
+    form: form
+
+# Generation Configuration
+  - name: guardrail_id
+    type: string
+    required: false
+    label:
+      en_US: Guardrail ID
+      zh_Hans: 防护栏ID
+    human_description:
+      en_US: ID of the guardrail to apply
+      zh_Hans: 要应用的防护栏ID
+    form: form
+
+  - name: guardrail_version
+    type: string
+    required: false
+    label:
+      en_US: Guardrail Version
+      zh_Hans: 防护栏版本
+    human_description:
+      en_US: Version of the guardrail to apply
+      zh_Hans: 要应用的防护栏版本
+    form: form
+
+  - name: max_tokens
+    type: number
+    required: false
+    label:
+      en_US: Maximum Tokens
+      zh_Hans: 最大令牌数
+    human_description:
+      en_US: Maximum number of tokens to generate
+      zh_Hans: 生成的最大令牌数
+    default: 2048
+    form: form
+
+  - name: stop_sequences
+    type: string
+    required: false
+    label:
+      en_US: Stop Sequences
+      zh_Hans: 停止序列
+    human_description:
+      en_US: JSON array of strings that will stop generation when encountered
+      zh_Hans: JSON数组格式的字符串，遇到这些序列时将停止生成
+    default: "[]"
+    form: form
+
+  - name: temperature
+    type: number
+    required: false
+    label:
+      en_US: Temperature
+      zh_Hans: 温度
+    human_description:
+      en_US: Controls randomness in the output (0-1)
+      zh_Hans: 控制输出的随机性（0-1）
+    default: 0.7
+    min: 0
+    max: 1
+    form: form
+
+  - name: top_p
+    type: number
+    required: false
+    label:
+      en_US: Top P
+      zh_Hans: Top P值
+    human_description:
+      en_US: Controls diversity via nucleus sampling (0-1)
+      zh_Hans: 通过核采样控制多样性（0-1）
+    default: 0.95
+    min: 0
+    max: 1
+    form: form
+
+  - name: performance_mode
+    type: select
+    required: false
+    label:
+      en_US: Performance Mode
+      zh_Hans: 性能模式
+    human_description:
+      en_US: Select performance optimization mode(performanceConfig.latency)
+      zh_Hans: 选择性能优化模式(performanceConfig.latency)
+    default: standard
+    options:
+      - value: standard
+        label:
+          en_US: Standard
+          zh_Hans: 标准
+      - value: optimized
+        label:
+          en_US: Optimized
+          zh_Hans: 优化
+    form: form
+
+  - name: prompt_template
+    type: string
+    required: false
+    label:
+      en_US: Prompt Template
+      zh_Hans: 提示模板
+    human_description:
+      en_US: Custom prompt template for generation
+      zh_Hans: 用于生成的自定义提示模板
+    form: form
+
+  - name: additional_model_fields
+    type: string
+    required: false
+    label:
+      en_US: Additional Model Fields
+      zh_Hans: 额外模型字段
+    human_description:
+      en_US: JSON formatted additional fields for model configuration
+      zh_Hans: JSON格式的额外模型配置字段
+    default: "{}"
+    form: form
+
+# Orchestration Configuration
+  - name: orchestration_max_tokens
+    type: number
+    required: false
+    label:
+      en_US: Orchestration Maximum Tokens
+      zh_Hans: 编排最大令牌数
+    human_description:
+      en_US: Maximum number of tokens for orchestration
+      zh_Hans: 编排过程的最大令牌数
+    default: 2048
+    form: form
+
+  - name: orchestration_stop_sequences
+    type: string
+    required: false
+    label:
+      en_US: Orchestration Stop Sequences
+      zh_Hans: 编排停止序列
+    human_description:
+      en_US: JSON array of strings that will stop orchestration when encountered
+      zh_Hans: JSON数组格式的字符串，遇到这些序列时将停止编排
+    default: "[]"
+    form: form
+
+  - name: orchestration_temperature
+    type: number
+    required: false
+    label:
+      en_US: Orchestration Temperature
+      zh_Hans: 编排温度
+    human_description:
+      en_US: Controls randomness in the orchestration output (0-1)
+      zh_Hans: 控制编排输出的随机性（0-1）
+    default: 0.7
+    min: 0
+    max: 1
+    form: form
+
+  - name: orchestration_top_p
+    type: number
+    required: false
+    label:
+      en_US: Orchestration Top P
+      zh_Hans: 编排Top P值
+    human_description:
+      en_US: Controls diversity via nucleus sampling in orchestration (0-1)
+      zh_Hans: 通过核采样控制编排的多样性（0-1）
+    default: 0.95
+    min: 0
+    max: 1
+    form: form
+
+  - name: orchestration_performance_mode
+    type: select
+    required: false
+    label:
+      en_US: Orchestration Performance Mode
+      zh_Hans: 编排性能模式
+    human_description:
+      en_US: Select performance optimization mode for orchestration
+      zh_Hans: 选择编排的性能优化模式
+    default: standard
+    options:
+      - value: standard
+        label:
+          en_US: Standard
+          zh_Hans: 标准
+      - value: optimized
+        label:
+          en_US: Optimized
+          zh_Hans: 优化
+    form: form
+
+  - name: orchestration_prompt_template
+    type: string
+    required: false
+    label:
+      en_US: Orchestration Prompt Template
+      zh_Hans: 编排提示模板
+    human_description:
+      en_US: Custom prompt template for orchestration
+      zh_Hans: 用于编排的自定义提示模板
+    form: form
+
+  - name: orchestration_additional_model_fields
+    type: string
+    required: false
+    label:
+      en_US: Orchestration Additional Model Fields
+      zh_Hans: 编排额外模型字段
+    human_description:
+      en_US: JSON formatted additional fields for orchestration model configuration
+      zh_Hans: JSON格式的编排模型额外配置字段
+    default: "{}"
+    form: form

api/core/tools/provider/builtin/gitee_ai/tools/risk-control.py

@@ -0,0 +1,26 @@
+from typing import Any, Union
+
+import requests
+
+from core.tools.entities.tool_entities import ToolInvokeMessage
+from core.tools.tool.builtin_tool import BuiltinTool
+
+
+class GiteeAIToolRiskControl(BuiltinTool):
+    def _invoke(
+        self, user_id: str, tool_parameters: dict[str, Any]
+    ) -> Union[ToolInvokeMessage, list[ToolInvokeMessage]]:
+        headers = {
+            "content-type": "application/json",
+            "authorization": f"Bearer {self.runtime.credentials['api_key']}",
+        }
+
+        inputs = [{"type": "text", "text": tool_parameters.get("input-text")}]
+        model = tool_parameters.get("model", "Security-semantic-filtering")
+        payload = {"model": model, "input": inputs}
+        url = "https://ai.gitee.com/v1/moderations"
+        response = requests.post(url, json=payload, headers=headers)
+        if response.status_code != 200:
+            return self.create_text_message(f"Got Error Response:{response.text}")
+
+        return [self.create_text_message(response.content.decode("utf-8"))]

api/core/tools/provider/builtin/gitee_ai/tools/risk-control.yaml

@@ -0,0 +1,32 @@
+identity:
+  name: risk control
+  author: gitee_ai
+  label:
+    en_US: risk control identification
+    zh_Hans: 风控识别
+  icon: icon.svg
+description:
+  human:
+    en_US: Ensuring the protection and compliance of sensitive information through the filtering and analysis of data semantics
+    zh_Hans: 通过对数据语义的过滤和分析，确保敏感信息的保护和合规性
+  llm: This tool is used to risk control identification.
+parameters:
+  - name: model
+    type: string
+    required: true
+    default: Security-semantic-filtering
+    label:
+      en_US: Service Model
+      zh_Hans: 服务模型
+    form: form
+  - name: input-text
+    type: string
+    required: true
+    label:
+      en_US: Input Text
+      zh_Hans: 输入文本
+    human_description:
+      en_US: The text input for filtering and analysis.
+      zh_Hans: 用于分析过滤的文本
+    llm_description: The text input for filtering and analysis.
+    form: llm

api/core/workflow/nodes/llm/node.py

@@ -185,6 +185,8 @@ class LLMNode(BaseNode[LLMNodeData]):
                     result_text = event.text
                     usage = event.usage
                     finish_reason = event.finish_reason
+                    # deduct quota
+                    self.deduct_llm_quota(tenant_id=self.tenant_id, model_instance=model_instance, usage=usage)
                     break
         except LLMNodeError as e:
             yield RunCompletedEvent(
@@ -240,17 +242,7 @@ class LLMNode(BaseNode[LLMNodeData]):
             user=self.user_id,
         )
 
-        # handle invoke result
-        generator = self._handle_invoke_result(invoke_result=invoke_result)
-
-        usage = LLMUsage.empty_usage()
-        for event in generator:
-            yield event
-            if isinstance(event, ModelInvokeCompletedEvent):
-                usage = event.usage
-
-        # deduct quota
-        self.deduct_llm_quota(tenant_id=self.tenant_id, model_instance=model_instance, usage=usage)
+        return self._handle_invoke_result(invoke_result=invoke_result)
 
     def _handle_invoke_result(self, invoke_result: LLMResult | Generator) -> Generator[NodeEvent, None, None]:
         if isinstance(invoke_result, LLMResult):

api/docker/entrypoint.sh

@@ -20,11 +20,11 @@ if [[ "${MODE}" == "worker" ]]; then
     CONCURRENCY_OPTION="-c ${CELERY_WORKER_AMOUNT:-1}"
   fi
 
-  exec celery -A app.celery worker -P ${CELERY_WORKER_CLASS:-gevent} $CONCURRENCY_OPTION --loglevel ${LOG_LEVEL} \
+  exec celery -A app.celery worker -P ${CELERY_WORKER_CLASS:-gevent} $CONCURRENCY_OPTION --loglevel ${LOG_LEVEL:-INFO} \
     -Q ${CELERY_QUEUES:-dataset,mail,ops_trace,app_deletion}
 
 elif [[ "${MODE}" == "beat" ]]; then
-  exec celery -A app.celery beat --loglevel ${LOG_LEVEL}
+  exec celery -A app.celery beat --loglevel ${LOG_LEVEL:-INFO}
 else
   if [[ "${DEBUG}" == "true" ]]; then
     exec flask run --host=${DIFY_BIND_ADDRESS:-0.0.0.0} --port=${DIFY_PORT:-5001} --debug

api/extensions/ext_logging.py

@@ -27,12 +27,11 @@ def init_app(app: DifyApp):
     # Always add StreamHandler to log to console
     sh = logging.StreamHandler(sys.stdout)
     sh.addFilter(RequestIdFilter())
-    log_formatter = logging.Formatter(fmt=dify_config.LOG_FORMAT)
-    sh.setFormatter(log_formatter)
     log_handlers.append(sh)
 
     logging.basicConfig(
         level=dify_config.LOG_LEVEL,
+        format=dify_config.LOG_FORMAT,
         datefmt=dify_config.LOG_DATEFORMAT,
         handlers=log_handlers,
         force=True,

api/models/model.py

@@ -1066,8 +1066,10 @@ class Message(db.Model):  # type: ignore[name-defined]
             "id": self.id,
             "app_id": self.app_id,
             "conversation_id": self.conversation_id,
+            "model_id": self.model_id,
             "inputs": self.inputs,
             "query": self.query,
+            "total_price": self.total_price,
             "message": self.message,
             "answer": self.answer,
             "status": self.status,
@@ -1088,7 +1090,9 @@ class Message(db.Model):  # type: ignore[name-defined]
             id=data["id"],
             app_id=data["app_id"],
             conversation_id=data["conversation_id"],
+            model_id=data["model_id"],
             inputs=data["inputs"],
+            total_price=data["total_price"],
             query=data["query"],
             message=data["message"],
             answer=data["answer"],

api/poetry.lock

@@ -922,7 +922,7 @@ version = "1.9.0"
 description = "Fast, simple object-to-object and broadcast signaling"
 optional = false
 python-versions = ">=3.9"
-groups = ["main", "tools"]
+groups = ["main", "dev", "tools"]
 markers = "python_version == \"3.11\" or python_version >= \"3.12\""
 files = [
     {file = "blinker-1.9.0-py3-none-any.whl", hash = "sha256:ba0efaa9080b619ff2f3459d1d500c57bddea4a6b424b60a91141db6fd2f08bc"},
@@ -931,36 +931,36 @@ files = [
 
 [[package]]
 name = "boto3"
-version = "1.35.74"
+version = "1.36.4"
 description = "The AWS SDK for Python"
 optional = false
 python-versions = ">=3.8"
 groups = ["main"]
 markers = "python_version == \"3.11\" or python_version >= \"3.12\""
 files = [
-    {file = "boto3-1.35.74-py3-none-any.whl", hash = "sha256:dab5bddbbe57dc707b6f6a1f25dc2823b8e234b6fe99fafef7fc406ab73031b9"},
-    {file = "boto3-1.35.74.tar.gz", hash = "sha256:88370c6845ba71a4dae7f6b357099df29b3965da584be040c8e72c9902bc9492"},
+    {file = "boto3-1.36.4-py3-none-any.whl", hash = "sha256:9f8f699e75ec63fcc98c4dd7290997c7c06c68d3ac8161ad4735fe71f5fe945c"},
+    {file = "boto3-1.36.4.tar.gz", hash = "sha256:eeceeb74ef8b65634d358c27aa074917f4449dc828f79301f1075232618eb502"},
 ]
 
 [package.dependencies]
-botocore = ">=1.35.74,<1.36.0"
+botocore = ">=1.36.4,<1.37.0"
 jmespath = ">=0.7.1,<2.0.0"
-s3transfer = ">=0.10.0,<0.11.0"
+s3transfer = ">=0.11.0,<0.12.0"
 
 [package.extras]
 crt = ["botocore[crt] (>=1.21.0,<2.0a0)"]
 
 [[package]]
 name = "botocore"
-version = "1.35.94"
+version = "1.36.5"
 description = "Low-level, data-driven core of boto 3."
 optional = false
 python-versions = ">=3.8"
 groups = ["main"]
 markers = "python_version == \"3.11\" or python_version >= \"3.12\""
 files = [
-    {file = "botocore-1.35.94-py3-none-any.whl", hash = "sha256:d784d944865d8279c79d2301fc09ac28b5221d4e7328fb4e23c642c253b9932c"},
-    {file = "botocore-1.35.94.tar.gz", hash = "sha256:2b3309b356541faa4d88bb957dcac1d8004aa44953c0b7d4521a6cc5d3d5d6ba"},
+    {file = "botocore-1.36.5-py3-none-any.whl", hash = "sha256:6d9f70afa9bf9d21407089dc22b8cc8ec6fa44866d4660858c062c74fc8555eb"},
+    {file = "botocore-1.36.5.tar.gz", hash = "sha256:234ed3d29a8954c37a551c933453bf14c6ae44a69a4f273ffef377a2612ca6a6"},
 ]
 
 [package.dependencies]
@@ -969,7 +969,7 @@ python-dateutil = ">=2.1,<3.0.0"
 urllib3 = {version = ">=1.25.4,<2.2.0 || >2.2.0,<3", markers = "python_version >= \"3.10\""}
 
 [package.extras]
-crt = ["awscrt (==0.22.0)"]
+crt = ["awscrt (==0.23.4)"]
 
 [[package]]
 name = "bottleneck"
@@ -1043,10 +1043,6 @@ files = [
     {file = "Brotli-1.1.0-cp310-cp310-musllinux_1_1_i686.whl", hash = "sha256:a37b8f0391212d29b3a91a799c8e4a2855e0576911cdfb2515487e30e322253d"},
     {file = "Brotli-1.1.0-cp310-cp310-musllinux_1_1_ppc64le.whl", hash = "sha256:e84799f09591700a4154154cab9787452925578841a94321d5ee8fb9a9a328f0"},
     {file = "Brotli-1.1.0-cp310-cp310-musllinux_1_1_x86_64.whl", hash = "sha256:f66b5337fa213f1da0d9000bc8dc0cb5b896b726eefd9c6046f699b169c41b9e"},
-    {file = "Brotli-1.1.0-cp310-cp310-musllinux_1_2_aarch64.whl", hash = "sha256:5dab0844f2cf82be357a0eb11a9087f70c5430b2c241493fc122bb6f2bb0917c"},
-    {file = "Brotli-1.1.0-cp310-cp310-musllinux_1_2_i686.whl", hash = "sha256:e4fe605b917c70283db7dfe5ada75e04561479075761a0b3866c081d035b01c1"},
-    {file = "Brotli-1.1.0-cp310-cp310-musllinux_1_2_ppc64le.whl", hash = "sha256:1e9a65b5736232e7a7f91ff3d02277f11d339bf34099a56cdab6a8b3410a02b2"},
-    {file = "Brotli-1.1.0-cp310-cp310-musllinux_1_2_x86_64.whl", hash = "sha256:58d4b711689366d4a03ac7957ab8c28890415e267f9b6589969e74b6e42225ec"},
     {file = "Brotli-1.1.0-cp310-cp310-win32.whl", hash = "sha256:be36e3d172dc816333f33520154d708a2657ea63762ec16b62ece02ab5e4daf2"},
     {file = "Brotli-1.1.0-cp310-cp310-win_amd64.whl", hash = "sha256:0c6244521dda65ea562d5a69b9a26120769b7a9fb3db2fe9545935ed6735b128"},
     {file = "Brotli-1.1.0-cp311-cp311-macosx_10_9_universal2.whl", hash = "sha256:a3daabb76a78f829cafc365531c972016e4aa8d5b4bf60660ad8ecee19df7ccc"},
@@ -1059,14 +1055,8 @@ files = [
     {file = "Brotli-1.1.0-cp311-cp311-musllinux_1_1_i686.whl", hash = "sha256:19c116e796420b0cee3da1ccec3b764ed2952ccfcc298b55a10e5610ad7885f9"},
     {file = "Brotli-1.1.0-cp311-cp311-musllinux_1_1_ppc64le.whl", hash = "sha256:510b5b1bfbe20e1a7b3baf5fed9e9451873559a976c1a78eebaa3b86c57b4265"},
     {file = "Brotli-1.1.0-cp311-cp311-musllinux_1_1_x86_64.whl", hash = "sha256:a1fd8a29719ccce974d523580987b7f8229aeace506952fa9ce1d53a033873c8"},
-    {file = "Brotli-1.1.0-cp311-cp311-musllinux_1_2_aarch64.whl", hash = "sha256:c247dd99d39e0338a604f8c2b3bc7061d5c2e9e2ac7ba9cc1be5a69cb6cd832f"},
-    {file = "Brotli-1.1.0-cp311-cp311-musllinux_1_2_i686.whl", hash = "sha256:1b2c248cd517c222d89e74669a4adfa5577e06ab68771a529060cf5a156e9757"},
-    {file = "Brotli-1.1.0-cp311-cp311-musllinux_1_2_ppc64le.whl", hash = "sha256:2a24c50840d89ded6c9a8fdc7b6ed3692ed4e86f1c4a4a938e1e92def92933e0"},
-    {file = "Brotli-1.1.0-cp311-cp311-musllinux_1_2_x86_64.whl", hash = "sha256:f31859074d57b4639318523d6ffdca586ace54271a73ad23ad021acd807eb14b"},
     {file = "Brotli-1.1.0-cp311-cp311-win32.whl", hash = "sha256:39da8adedf6942d76dc3e46653e52df937a3c4d6d18fdc94a7c29d263b1f5b50"},
     {file = "Brotli-1.1.0-cp311-cp311-win_amd64.whl", hash = "sha256:aac0411d20e345dc0920bdec5548e438e999ff68d77564d5e9463a7ca9d3e7b1"},
-    {file = "Brotli-1.1.0-cp312-cp312-macosx_10_13_universal2.whl", hash = "sha256:32d95b80260d79926f5fab3c41701dbb818fde1c9da590e77e571eefd14abe28"},
-    {file = "Brotli-1.1.0-cp312-cp312-macosx_10_13_x86_64.whl", hash = "sha256:b760c65308ff1e462f65d69c12e4ae085cff3b332d894637f6273a12a482d09f"},
     {file = "Brotli-1.1.0-cp312-cp312-macosx_10_9_universal2.whl", hash = "sha256:316cc9b17edf613ac76b1f1f305d2a748f1b976b033b049a6ecdfd5612c70409"},
     {file = "Brotli-1.1.0-cp312-cp312-macosx_10_9_x86_64.whl", hash = "sha256:caf9ee9a5775f3111642d33b86237b05808dafcd6268faa492250e9b78046eb2"},
     {file = "Brotli-1.1.0-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:70051525001750221daa10907c77830bc889cb6d865cc0b813d9db7fefc21451"},
@@ -1077,24 +1067,8 @@ files = [
     {file = "Brotli-1.1.0-cp312-cp312-musllinux_1_1_i686.whl", hash = "sha256:4093c631e96fdd49e0377a9c167bfd75b6d0bad2ace734c6eb20b348bc3ea180"},
     {file = "Brotli-1.1.0-cp312-cp312-musllinux_1_1_ppc64le.whl", hash = "sha256:7e4c4629ddad63006efa0ef968c8e4751c5868ff0b1c5c40f76524e894c50248"},
     {file = "Brotli-1.1.0-cp312-cp312-musllinux_1_1_x86_64.whl", hash = "sha256:861bf317735688269936f755fa136a99d1ed526883859f86e41a5d43c61d8966"},
-    {file = "Brotli-1.1.0-cp312-cp312-musllinux_1_2_aarch64.whl", hash = "sha256:87a3044c3a35055527ac75e419dfa9f4f3667a1e887ee80360589eb8c90aabb9"},
-    {file = "Brotli-1.1.0-cp312-cp312-musllinux_1_2_i686.whl", hash = "sha256:c5529b34c1c9d937168297f2c1fde7ebe9ebdd5e121297ff9c043bdb2ae3d6fb"},
-    {file = "Brotli-1.1.0-cp312-cp312-musllinux_1_2_ppc64le.whl", hash = "sha256:ca63e1890ede90b2e4454f9a65135a4d387a4585ff8282bb72964fab893f2111"},
-    {file = "Brotli-1.1.0-cp312-cp312-musllinux_1_2_x86_64.whl", hash = "sha256:e79e6520141d792237c70bcd7a3b122d00f2613769ae0cb61c52e89fd3443839"},
     {file = "Brotli-1.1.0-cp312-cp312-win32.whl", hash = "sha256:5f4d5ea15c9382135076d2fb28dde923352fe02951e66935a9efaac8f10e81b0"},
     {file = "Brotli-1.1.0-cp312-cp312-win_amd64.whl", hash = "sha256:906bc3a79de8c4ae5b86d3d75a8b77e44404b0f4261714306e3ad248d8ab0951"},
-    {file = "Brotli-1.1.0-cp313-cp313-macosx_10_13_universal2.whl", hash = "sha256:8bf32b98b75c13ec7cf774164172683d6e7891088f6316e54425fde1efc276d5"},
-    {file = "Brotli-1.1.0-cp313-cp313-macosx_10_13_x86_64.whl", hash = "sha256:7bc37c4d6b87fb1017ea28c9508b36bbcb0c3d18b4260fcdf08b200c74a6aee8"},
-    {file = "Brotli-1.1.0-cp313-cp313-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:3c0ef38c7a7014ffac184db9e04debe495d317cc9c6fb10071f7fefd93100a4f"},
-    {file = "Brotli-1.1.0-cp313-cp313-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:91d7cc2a76b5567591d12c01f019dd7afce6ba8cba6571187e21e2fc418ae648"},
-    {file = "Brotli-1.1.0-cp313-cp313-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:a93dde851926f4f2678e704fadeb39e16c35d8baebd5252c9fd94ce8ce68c4a0"},
-    {file = "Brotli-1.1.0-cp313-cp313-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:f0db75f47be8b8abc8d9e31bc7aad0547ca26f24a54e6fd10231d623f183d089"},
-    {file = "Brotli-1.1.0-cp313-cp313-musllinux_1_2_aarch64.whl", hash = "sha256:6967ced6730aed543b8673008b5a391c3b1076d834ca438bbd70635c73775368"},
-    {file = "Brotli-1.1.0-cp313-cp313-musllinux_1_2_i686.whl", hash = "sha256:7eedaa5d036d9336c95915035fb57422054014ebdeb6f3b42eac809928e40d0c"},
-    {file = "Brotli-1.1.0-cp313-cp313-musllinux_1_2_ppc64le.whl", hash = "sha256:d487f5432bf35b60ed625d7e1b448e2dc855422e87469e3f450aa5552b0eb284"},
-    {file = "Brotli-1.1.0-cp313-cp313-musllinux_1_2_x86_64.whl", hash = "sha256:832436e59afb93e1836081a20f324cb185836c617659b07b129141a8426973c7"},
-    {file = "Brotli-1.1.0-cp313-cp313-win32.whl", hash = "sha256:43395e90523f9c23a3d5bdf004733246fba087f2948f87ab28015f12359ca6a0"},
-    {file = "Brotli-1.1.0-cp313-cp313-win_amd64.whl", hash = "sha256:9011560a466d2eb3f5a6e4929cf4a09be405c64154e12df0dd72713f6500e32b"},
     {file = "Brotli-1.1.0-cp36-cp36m-macosx_10_9_x86_64.whl", hash = "sha256:a090ca607cbb6a34b0391776f0cb48062081f5f60ddcce5d11838e67a01928d1"},
     {file = "Brotli-1.1.0-cp36-cp36m-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:2de9d02f5bda03d27ede52e8cfe7b865b066fa49258cbab568720aa5be80a47d"},
     {file = "Brotli-1.1.0-cp36-cp36m-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:2333e30a5e00fe0fe55903c8832e08ee9c3b1382aacf4db26664a16528d51b4b"},
@@ -1104,10 +1078,6 @@ files = [
     {file = "Brotli-1.1.0-cp36-cp36m-musllinux_1_1_i686.whl", hash = "sha256:fd5f17ff8f14003595ab414e45fce13d073e0762394f957182e69035c9f3d7c2"},
     {file = "Brotli-1.1.0-cp36-cp36m-musllinux_1_1_ppc64le.whl", hash = "sha256:069a121ac97412d1fe506da790b3e69f52254b9df4eb665cd42460c837193354"},
     {file = "Brotli-1.1.0-cp36-cp36m-musllinux_1_1_x86_64.whl", hash = "sha256:e93dfc1a1165e385cc8239fab7c036fb2cd8093728cbd85097b284d7b99249a2"},
-    {file = "Brotli-1.1.0-cp36-cp36m-musllinux_1_2_aarch64.whl", hash = "sha256:aea440a510e14e818e67bfc4027880e2fb500c2ccb20ab21c7a7c8b5b4703d75"},
-    {file = "Brotli-1.1.0-cp36-cp36m-musllinux_1_2_i686.whl", hash = "sha256:6974f52a02321b36847cd19d1b8e381bf39939c21efd6ee2fc13a28b0d99348c"},
-    {file = "Brotli-1.1.0-cp36-cp36m-musllinux_1_2_ppc64le.whl", hash = "sha256:a7e53012d2853a07a4a79c00643832161a910674a893d296c9f1259859a289d2"},
-    {file = "Brotli-1.1.0-cp36-cp36m-musllinux_1_2_x86_64.whl", hash = "sha256:d7702622a8b40c49bffb46e1e3ba2e81268d5c04a34f460978c6b5517a34dd52"},
     {file = "Brotli-1.1.0-cp36-cp36m-win32.whl", hash = "sha256:a599669fd7c47233438a56936988a2478685e74854088ef5293802123b5b2460"},
     {file = "Brotli-1.1.0-cp36-cp36m-win_amd64.whl", hash = "sha256:d143fd47fad1db3d7c27a1b1d66162e855b5d50a89666af46e1679c496e8e579"},
     {file = "Brotli-1.1.0-cp37-cp37m-macosx_10_9_x86_64.whl", hash = "sha256:11d00ed0a83fa22d29bc6b64ef636c4552ebafcef57154b4ddd132f5638fbd1c"},
@@ -1119,10 +1089,6 @@ files = [
     {file = "Brotli-1.1.0-cp37-cp37m-musllinux_1_1_i686.whl", hash = "sha256:919e32f147ae93a09fe064d77d5ebf4e35502a8df75c29fb05788528e330fe74"},
     {file = "Brotli-1.1.0-cp37-cp37m-musllinux_1_1_ppc64le.whl", hash = "sha256:23032ae55523cc7bccb4f6a0bf368cd25ad9bcdcc1990b64a647e7bbcce9cb5b"},
     {file = "Brotli-1.1.0-cp37-cp37m-musllinux_1_1_x86_64.whl", hash = "sha256:224e57f6eac61cc449f498cc5f0e1725ba2071a3d4f48d5d9dffba42db196438"},
-    {file = "Brotli-1.1.0-cp37-cp37m-musllinux_1_2_aarch64.whl", hash = "sha256:cb1dac1770878ade83f2ccdf7d25e494f05c9165f5246b46a621cc849341dc01"},
-    {file = "Brotli-1.1.0-cp37-cp37m-musllinux_1_2_i686.whl", hash = "sha256:3ee8a80d67a4334482d9712b8e83ca6b1d9bc7e351931252ebef5d8f7335a547"},
-    {file = "Brotli-1.1.0-cp37-cp37m-musllinux_1_2_ppc64le.whl", hash = "sha256:5e55da2c8724191e5b557f8e18943b1b4839b8efc3ef60d65985bcf6f587dd38"},
-    {file = "Brotli-1.1.0-cp37-cp37m-musllinux_1_2_x86_64.whl", hash = "sha256:d342778ef319e1026af243ed0a07c97acf3bad33b9f29e7ae6a1f68fd083e90c"},
     {file = "Brotli-1.1.0-cp37-cp37m-win32.whl", hash = "sha256:587ca6d3cef6e4e868102672d3bd9dc9698c309ba56d41c2b9c85bbb903cdb95"},
     {file = "Brotli-1.1.0-cp37-cp37m-win_amd64.whl", hash = "sha256:2954c1c23f81c2eaf0b0717d9380bd348578a94161a65b3a2afc62c86467dd68"},
     {file = "Brotli-1.1.0-cp38-cp38-macosx_10_9_universal2.whl", hash = "sha256:efa8b278894b14d6da122a72fefcebc28445f2d3f880ac59d46c90f4c13be9a3"},
@@ -1135,10 +1101,6 @@ files = [
     {file = "Brotli-1.1.0-cp38-cp38-musllinux_1_1_i686.whl", hash = "sha256:1ab4fbee0b2d9098c74f3057b2bc055a8bd92ccf02f65944a241b4349229185a"},
     {file = "Brotli-1.1.0-cp38-cp38-musllinux_1_1_ppc64le.whl", hash = "sha256:141bd4d93984070e097521ed07e2575b46f817d08f9fa42b16b9b5f27b5ac088"},
     {file = "Brotli-1.1.0-cp38-cp38-musllinux_1_1_x86_64.whl", hash = "sha256:fce1473f3ccc4187f75b4690cfc922628aed4d3dd013d047f95a9b3919a86596"},
-    {file = "Brotli-1.1.0-cp38-cp38-musllinux_1_2_aarch64.whl", hash = "sha256:d2b35ca2c7f81d173d2fadc2f4f31e88cc5f7a39ae5b6db5513cf3383b0e0ec7"},
-    {file = "Brotli-1.1.0-cp38-cp38-musllinux_1_2_i686.whl", hash = "sha256:af6fa6817889314555aede9a919612b23739395ce767fe7fcbea9a80bf140fe5"},
-    {file = "Brotli-1.1.0-cp38-cp38-musllinux_1_2_ppc64le.whl", hash = "sha256:2feb1d960f760a575dbc5ab3b1c00504b24caaf6986e2dc2b01c09c87866a943"},
-    {file = "Brotli-1.1.0-cp38-cp38-musllinux_1_2_x86_64.whl", hash = "sha256:4410f84b33374409552ac9b6903507cdb31cd30d2501fc5ca13d18f73548444a"},
     {file = "Brotli-1.1.0-cp38-cp38-win32.whl", hash = "sha256:db85ecf4e609a48f4b29055f1e144231b90edc90af7481aa731ba2d059226b1b"},
     {file = "Brotli-1.1.0-cp38-cp38-win_amd64.whl", hash = "sha256:3d7954194c36e304e1523f55d7042c59dc53ec20dd4e9ea9d151f1b62b4415c0"},
     {file = "Brotli-1.1.0-cp39-cp39-macosx_10_9_universal2.whl", hash = "sha256:5fb2ce4b8045c78ebbc7b8f3c15062e435d47e7393cc57c25115cfd49883747a"},
@@ -1151,10 +1113,6 @@ files = [
     {file = "Brotli-1.1.0-cp39-cp39-musllinux_1_1_i686.whl", hash = "sha256:949f3b7c29912693cee0afcf09acd6ebc04c57af949d9bf77d6101ebb61e388c"},
     {file = "Brotli-1.1.0-cp39-cp39-musllinux_1_1_ppc64le.whl", hash = "sha256:89f4988c7203739d48c6f806f1e87a1d96e0806d44f0fba61dba81392c9e474d"},
     {file = "Brotli-1.1.0-cp39-cp39-musllinux_1_1_x86_64.whl", hash = "sha256:de6551e370ef19f8de1807d0a9aa2cdfdce2e85ce88b122fe9f6b2b076837e59"},
-    {file = "Brotli-1.1.0-cp39-cp39-musllinux_1_2_aarch64.whl", hash = "sha256:0737ddb3068957cf1b054899b0883830bb1fec522ec76b1098f9b6e0f02d9419"},
-    {file = "Brotli-1.1.0-cp39-cp39-musllinux_1_2_i686.whl", hash = "sha256:4f3607b129417e111e30637af1b56f24f7a49e64763253bbc275c75fa887d4b2"},
-    {file = "Brotli-1.1.0-cp39-cp39-musllinux_1_2_ppc64le.whl", hash = "sha256:6c6e0c425f22c1c719c42670d561ad682f7bfeeef918edea971a79ac5252437f"},
-    {file = "Brotli-1.1.0-cp39-cp39-musllinux_1_2_x86_64.whl", hash = "sha256:494994f807ba0b92092a163a0a283961369a65f6cbe01e8891132b7a320e61eb"},
     {file = "Brotli-1.1.0-cp39-cp39-win32.whl", hash = "sha256:f0d8a7a6b5983c2496e364b969f0e526647a06b075d034f3297dc66f3b360c64"},
     {file = "Brotli-1.1.0-cp39-cp39-win_amd64.whl", hash = "sha256:cdad5b9014d83ca68c25d2e9444e28e967ef16e80f6b436918c700c117a85467"},
     {file = "Brotli-1.1.0.tar.gz", hash = "sha256:81de08ac11bcb85841e440c13611c00b67d3bf82698314928d0b676362546724"},
@@ -1628,7 +1586,7 @@ version = "8.1.8"
 description = "Composable command line interface toolkit"
 optional = false
 python-versions = ">=3.7"
-groups = ["main", "lint", "tools", "vdb"]
+groups = ["main", "dev", "lint", "tools", "vdb"]
 markers = "python_version == \"3.11\" or python_version >= \"3.12\""
 files = [
     {file = "click-8.1.8-py3-none-any.whl", hash = "sha256:63c132bbbed01578a06712a2d1f497bb62d9c1c0d329b7903a866228027263b2"},
@@ -1868,7 +1826,7 @@ files = [
     {file = "colorama-0.4.6-py2.py3-none-any.whl", hash = "sha256:4f1d9991f5acc0ca119f9d443620b77f9d6b33703e51011c16baf57afb285fc6"},
     {file = "colorama-0.4.6.tar.gz", hash = "sha256:08695f5cb7ed6e0531a20572697297273c47b8cae5a63ffc6d6ed5c201be6e44"},
 ]
-markers = {main = "python_version == \"3.11\" or python_version >= \"3.12\"", dev = "(python_version == \"3.11\" or python_version >= \"3.12\") and sys_platform == \"win32\"", lint = "(python_version == \"3.11\" or python_version >= \"3.12\") and platform_system == \"Windows\"", tools = "(python_version == \"3.11\" or python_version >= \"3.12\") and platform_system == \"Windows\"", vdb = "(python_version == \"3.11\" or python_version >= \"3.12\") and (platform_system == \"Windows\" or os_name == \"nt\" or sys_platform == \"win32\")"}
+markers = {main = "python_version == \"3.11\" or python_version >= \"3.12\"", dev = "(python_version == \"3.11\" or python_version >= \"3.12\") and (platform_system == \"Windows\" or sys_platform == \"win32\")", lint = "(python_version == \"3.11\" or python_version >= \"3.12\") and platform_system == \"Windows\"", tools = "(python_version == \"3.11\" or python_version >= \"3.12\") and platform_system == \"Windows\"", vdb = "(python_version == \"3.11\" or python_version >= \"3.12\") and (platform_system == \"Windows\" or os_name == \"nt\" or sys_platform == \"win32\")"}
 
 [[package]]
 name = "coloredlogs"
@@ -2834,7 +2792,7 @@ version = "3.1.0"
 description = "A simple framework for building complex web applications."
 optional = false
 python-versions = ">=3.9"
-groups = ["main", "tools"]
+groups = ["main", "dev", "tools"]
 markers = "python_version == \"3.11\" or python_version >= \"3.12\""
 files = [
     {file = "flask-3.1.0-py3-none-any.whl", hash = "sha256:d667207822eb83f1c4b50949b1623c8fc8d51f2341d65f72e1a1815397551136"},
@@ -2973,7 +2931,7 @@ version = "3.1.1"
 description = "Add SQLAlchemy support to your Flask application."
 optional = false
 python-versions = ">=3.8"
-groups = ["main"]
+groups = ["main", "dev"]
 markers = "python_version == \"3.11\" or python_version >= \"3.12\""
 files = [
     {file = "flask_sqlalchemy-3.1.1-py3-none-any.whl", hash = "sha256:4ba4be7f419dc72f4efd8802d69974803c37259dd42f3913b0dcf75c9447e0a0"},
@@ -3836,7 +3794,7 @@ version = "3.1.1"
 description = "Lightweight in-process concurrent programming"
 optional = false
 python-versions = ">=3.7"
-groups = ["main", "tools", "vdb"]
+groups = ["main", "dev", "tools", "vdb"]
 files = [
     {file = "greenlet-3.1.1-cp310-cp310-macosx_11_0_universal2.whl", hash = "sha256:0bbae94a29c9e5c7e4a2b7f0aae5c17e8e90acbfd3bf6270eeba60c39fce3563"},
     {file = "greenlet-3.1.1-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:0fde093fb93f35ca72a556cf72c92ea3ebfda3d79fc35bb19fbe685853869a83"},
@@ -3912,7 +3870,7 @@ files = [
     {file = "greenlet-3.1.1-cp39-cp39-win_amd64.whl", hash = "sha256:3319aa75e0e0639bc15ff54ca327e8dc7a6fe404003496e3c6925cd3142e0e22"},
     {file = "greenlet-3.1.1.tar.gz", hash = "sha256:4ce3ac6cdb6adf7946475d7ef31777c26d94bccc377e070a7986bd2d5c515467"},
 ]
-markers = {main = "(python_version == \"3.11\" or python_version >= \"3.12\") and (platform_machine == \"aarch64\" or platform_machine == \"ppc64le\" or platform_machine == \"x86_64\" or platform_machine == \"amd64\" or platform_machine == \"AMD64\" or platform_machine == \"win32\" or platform_machine == \"WIN32\" or platform_python_implementation == \"CPython\")", tools = "(python_version == \"3.11\" or python_version >= \"3.12\") and (platform_machine == \"aarch64\" or platform_machine == \"ppc64le\" or platform_machine == \"x86_64\" or platform_machine == \"amd64\" or platform_machine == \"AMD64\" or platform_machine == \"win32\" or platform_machine == \"WIN32\")", vdb = "(python_version == \"3.11\" or python_version >= \"3.12\") and (platform_machine == \"aarch64\" or platform_machine == \"ppc64le\" or platform_machine == \"x86_64\" or platform_machine == \"amd64\" or platform_machine == \"AMD64\" or platform_machine == \"win32\" or platform_machine == \"WIN32\")"}
+markers = {main = "(python_version == \"3.11\" or python_version >= \"3.12\") and (platform_machine == \"aarch64\" or platform_machine == \"ppc64le\" or platform_machine == \"x86_64\" or platform_machine == \"amd64\" or platform_machine == \"AMD64\" or platform_machine == \"win32\" or platform_machine == \"WIN32\" or platform_python_implementation == \"CPython\")", dev = "(python_version == \"3.11\" or python_version >= \"3.12\") and (platform_machine == \"aarch64\" or platform_machine == \"ppc64le\" or platform_machine == \"x86_64\" or platform_machine == \"amd64\" or platform_machine == \"AMD64\" or platform_machine == \"win32\" or platform_machine == \"WIN32\")", tools = "(python_version == \"3.11\" or python_version >= \"3.12\") and (platform_machine == \"aarch64\" or platform_machine == \"ppc64le\" or platform_machine == \"x86_64\" or platform_machine == \"amd64\" or platform_machine == \"AMD64\" or platform_machine == \"win32\" or platform_machine == \"WIN32\")", vdb = "(python_version == \"3.11\" or python_version >= \"3.12\") and (platform_machine == \"aarch64\" or platform_machine == \"ppc64le\" or platform_machine == \"x86_64\" or platform_machine == \"amd64\" or platform_machine == \"AMD64\" or platform_machine == \"win32\" or platform_machine == \"WIN32\")"}
 
 [package.extras]
 docs = ["Sphinx", "furo"]
@@ -4587,7 +4545,7 @@ version = "2.2.0"
 description = "Safely pass data to untrusted environments and back."
 optional = false
 python-versions = ">=3.8"
-groups = ["main", "tools"]
+groups = ["main", "dev", "tools"]
 markers = "python_version == \"3.11\" or python_version >= \"3.12\""
 files = [
     {file = "itsdangerous-2.2.0-py3-none-any.whl", hash = "sha256:c6242fc49e35958c8b15141343aa660db5fc54d4f13a1db01a3f5891b98700ef"},
@@ -4624,7 +4582,7 @@ version = "3.1.5"
 description = "A very fast and expressive template engine."
 optional = false
 python-versions = ">=3.7"
-groups = ["main", "tools"]
+groups = ["main", "dev", "tools"]
 markers = "python_version == \"3.11\" or python_version >= \"3.12\""
 files = [
     {file = "jinja2-3.1.5-py3-none-any.whl", hash = "sha256:aba0f4dc9ed8013c424088f68a5c226f7d6097ed89b246d7749c2ec4175c6adb"},
@@ -5556,7 +5514,7 @@ version = "3.0.2"
 description = "Safely add untrusted strings to HTML/XML markup."
 optional = false
 python-versions = ">=3.9"
-groups = ["main", "tools"]
+groups = ["main", "dev", "tools"]
 markers = "python_version == \"3.11\" or python_version >= \"3.12\""
 files = [
     {file = "MarkupSafe-3.0.2-cp310-cp310-macosx_10_9_universal2.whl", hash = "sha256:7e94c425039cde14257288fd61dcfb01963e658efbc0ff54f5306b06054700f8"},
@@ -9563,22 +9521,22 @@ files = [
 
 [[package]]
 name = "s3transfer"
-version = "0.10.4"
+version = "0.11.2"
 description = "An Amazon S3 Transfer Manager"
 optional = false
 python-versions = ">=3.8"
 groups = ["main"]
 markers = "python_version == \"3.11\" or python_version >= \"3.12\""
 files = [
-    {file = "s3transfer-0.10.4-py3-none-any.whl", hash = "sha256:244a76a24355363a68164241438de1b72f8781664920260c48465896b712a41e"},
-    {file = "s3transfer-0.10.4.tar.gz", hash = "sha256:29edc09801743c21eb5ecbc617a152df41d3c287f67b615f73e5f750583666a7"},
+    {file = "s3transfer-0.11.2-py3-none-any.whl", hash = "sha256:be6ecb39fadd986ef1701097771f87e4d2f821f27f6071c872143884d2950fbc"},
+    {file = "s3transfer-0.11.2.tar.gz", hash = "sha256:3b39185cb72f5acc77db1a58b6e25b977f28d20496b6e58d6813d75f464d632f"},
 ]
 
 [package.dependencies]
-botocore = ">=1.33.2,<2.0a.0"
+botocore = ">=1.36.0,<2.0a.0"
 
 [package.extras]
-crt = ["botocore[crt] (>=1.33.2,<2.0a.0)"]
+crt = ["botocore[crt] (>=1.36.0,<2.0a.0)"]
 
 [[package]]
 name = "safetensors"
@@ -10159,7 +10117,7 @@ version = "2.0.35"
 description = "Database Abstraction Library"
 optional = false
 python-versions = ">=3.7"
-groups = ["main", "tools", "vdb"]
+groups = ["main", "dev", "tools", "vdb"]
 markers = "python_version == \"3.11\" or python_version >= \"3.12\""
 files = [
     {file = "SQLAlchemy-2.0.35-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:67219632be22f14750f0d1c70e62f204ba69d28f62fd6432ba05ab295853de9b"},
@@ -10888,26 +10846,179 @@ rich = ">=10.11.0"
 shellingham = ">=1.3.0"
 typing-extensions = ">=3.7.4.3"
 
+[[package]]
+name = "types-beautifulsoup4"
+version = "4.12.0.20241020"
+description = "Typing stubs for beautifulsoup4"
+optional = false
+python-versions = ">=3.8"
+groups = ["dev"]
+markers = "python_version == \"3.11\" or python_version >= \"3.12\""
+files = [
+    {file = "types-beautifulsoup4-4.12.0.20241020.tar.gz", hash = "sha256:158370d08d0cd448bd11b132a50ff5279237a5d4b5837beba074de152a513059"},
+    {file = "types_beautifulsoup4-4.12.0.20241020-py3-none-any.whl", hash = "sha256:c95e66ce15a4f5f0835f7fbc5cd886321ae8294f977c495424eaf4225307fd30"},
+]
+
+[package.dependencies]
+types-html5lib = "*"
+
+[[package]]
+name = "types-flask-cors"
+version = "5.0.0.20240902"
+description = "Typing stubs for Flask-Cors"
+optional = false
+python-versions = ">=3.8"
+groups = ["dev"]
+markers = "python_version == \"3.11\" or python_version >= \"3.12\""
+files = [
+    {file = "types-Flask-Cors-5.0.0.20240902.tar.gz", hash = "sha256:8921b273bf7cd9636df136b66408efcfa6338a935e5c8f53f5eff1cee03f3394"},
+    {file = "types_Flask_Cors-5.0.0.20240902-py3-none-any.whl", hash = "sha256:595e5f36056cd128ab905832e055f2e5d116fbdc685356eea4490bc77df82137"},
+]
+
+[package.dependencies]
+Flask = ">=2.0.0"
+
+[[package]]
+name = "types-flask-migrate"
+version = "4.1.0.20250112"
+description = "Typing stubs for Flask-Migrate"
+optional = false
+python-versions = ">=3.9"
+groups = ["dev"]
+markers = "python_version == \"3.11\" or python_version >= \"3.12\""
+files = [
+    {file = "types_Flask_Migrate-4.1.0.20250112-py3-none-any.whl", hash = "sha256:1814fffc609c2ead784affd011de92f0beecd48044963a8c898dd107dc1b5969"},
+    {file = "types_flask_migrate-4.1.0.20250112.tar.gz", hash = "sha256:f2d2c966378ae7bb0660ec810e9af0a56ca03108235364c2a7b5e90418b0ff67"},
+]
+
+[package.dependencies]
+Flask = ">=2.0.0"
+Flask-SQLAlchemy = ">=3.0.1"
+
+[[package]]
+name = "types-html5lib"
+version = "1.1.11.20241018"
+description = "Typing stubs for html5lib"
+optional = false
+python-versions = ">=3.8"
+groups = ["dev"]
+markers = "python_version == \"3.11\" or python_version >= \"3.12\""
+files = [
+    {file = "types-html5lib-1.1.11.20241018.tar.gz", hash = "sha256:98042555ff78d9e3a51c77c918b1041acbb7eb6c405408d8a9e150ff5beccafa"},
+    {file = "types_html5lib-1.1.11.20241018-py3-none-any.whl", hash = "sha256:3f1e064d9ed2c289001ae6392c84c93833abb0816165c6ff0abfc304a779f403"},
+]
+
+[[package]]
+name = "types-openpyxl"
+version = "3.1.5.20241225"
+description = "Typing stubs for openpyxl"
+optional = false
+python-versions = ">=3.8"
+groups = ["dev"]
+markers = "python_version == \"3.11\" or python_version >= \"3.12\""
+files = [
+    {file = "types_openpyxl-3.1.5.20241225-py3-none-any.whl", hash = "sha256:903d92f58f42135b0614d609868c619aee12e1c7b65ccf8472dfd2706bcc6f47"},
+    {file = "types_openpyxl-3.1.5.20241225.tar.gz", hash = "sha256:3c076f4c6f114e1859b6857ffd486e96c938c0434451c60dc54c2bcb62750d78"},
+]
+
+[[package]]
+name = "types-protobuf"
+version = "5.29.1.20241207"
+description = "Typing stubs for protobuf"
+optional = false
+python-versions = ">=3.8"
+groups = ["dev"]
+markers = "python_version == \"3.11\" or python_version >= \"3.12\""
+files = [
+    {file = "types_protobuf-5.29.1.20241207-py3-none-any.whl", hash = "sha256:92893c42083e9b718c678badc0af7a9a1307b92afe1599e5cba5f3d35b668b2f"},
+    {file = "types_protobuf-5.29.1.20241207.tar.gz", hash = "sha256:2ebcadb8ab3ef2e3e2f067e0882906d64ba0dc65fc5b0fd7a8b692315b4a0be9"},
+]
+
+[[package]]
+name = "types-psutil"
+version = "6.1.0.20241221"
+description = "Typing stubs for psutil"
+optional = false
+python-versions = ">=3.8"
+groups = ["dev"]
+markers = "python_version == \"3.11\" or python_version >= \"3.12\""
+files = [
+    {file = "types_psutil-6.1.0.20241221-py3-none-any.whl", hash = "sha256:8498dbe13285a9ba7d4b2fa934c569cc380efc74e3dacdb34ae16d2cdf389ec3"},
+    {file = "types_psutil-6.1.0.20241221.tar.gz", hash = "sha256:600f5a36bd5e0eb8887f0e3f3ff2cf154d90690ad8123c8a707bba4ab94d3185"},
+]
+
+[[package]]
+name = "types-psycopg2"
+version = "2.9.21.20250121"
+description = "Typing stubs for psycopg2"
+optional = false
+python-versions = ">=3.9"
+groups = ["dev"]
+markers = "python_version == \"3.11\" or python_version >= \"3.12\""
+files = [
+    {file = "types_psycopg2-2.9.21.20250121-py3-none-any.whl", hash = "sha256:b890dc6f5a08b6433f0ff73a4ec9a834deedad3e914f2a4a6fd43df021f745f1"},
+    {file = "types_psycopg2-2.9.21.20250121.tar.gz", hash = "sha256:2b0e2cd0f3747af1ae25a7027898716d80209604770ef3cbf350fe055b9c349b"},
+]
+
+[[package]]
+name = "types-python-dateutil"
+version = "2.9.0.20241206"
+description = "Typing stubs for python-dateutil"
+optional = false
+python-versions = ">=3.8"
+groups = ["dev"]
+markers = "python_version == \"3.11\" or python_version >= \"3.12\""
+files = [
+    {file = "types_python_dateutil-2.9.0.20241206-py3-none-any.whl", hash = "sha256:e248a4bc70a486d3e3ec84d0dc30eec3a5f979d6e7ee4123ae043eedbb987f53"},
+    {file = "types_python_dateutil-2.9.0.20241206.tar.gz", hash = "sha256:18f493414c26ffba692a72369fea7a154c502646301ebfe3d56a04b3767284cb"},
+]
+
 [[package]]
 name = "types-pytz"
 version = "2024.2.0.20241221"
 description = "Typing stubs for pytz"
 optional = false
 python-versions = ">=3.8"
-groups = ["main"]
+groups = ["main", "dev"]
 markers = "python_version == \"3.11\" or python_version >= \"3.12\""
 files = [
     {file = "types_pytz-2024.2.0.20241221-py3-none-any.whl", hash = "sha256:8fc03195329c43637ed4f593663df721fef919b60a969066e22606edf0b53ad5"},
     {file = "types_pytz-2024.2.0.20241221.tar.gz", hash = "sha256:06d7cde9613e9f7504766a0554a270c369434b50e00975b3a4a0f6eed0f2c1a9"},
 ]
 
+[[package]]
+name = "types-pyyaml"
+version = "6.0.12.20241230"
+description = "Typing stubs for PyYAML"
+optional = false
+python-versions = ">=3.8"
+groups = ["dev"]
+markers = "python_version == \"3.11\" or python_version >= \"3.12\""
+files = [
+    {file = "types_PyYAML-6.0.12.20241230-py3-none-any.whl", hash = "sha256:fa4d32565219b68e6dee5f67534c722e53c00d1cfc09c435ef04d7353e1e96e6"},
+    {file = "types_pyyaml-6.0.12.20241230.tar.gz", hash = "sha256:7f07622dbd34bb9c8b264fe860a17e0efcad00d50b5f27e93984909d9363498c"},
+]
+
+[[package]]
+name = "types-regex"
+version = "2024.11.6.20241221"
+description = "Typing stubs for regex"
+optional = false
+python-versions = ">=3.8"
+groups = ["dev"]
+markers = "python_version == \"3.11\" or python_version >= \"3.12\""
+files = [
+    {file = "types_regex-2024.11.6.20241221-py3-none-any.whl", hash = "sha256:9d29ab639df22a86e15e2cc037e92ad100a4e8f4ecd2ad261d6f0c6d8d87f54e"},
+    {file = "types_regex-2024.11.6.20241221.tar.gz", hash = "sha256:903c7b557d935363ba01f07a75981c78ada7df66623e415f32bda2afecfa5cca"},
+]
+
 [[package]]
 name = "types-requests"
 version = "2.32.0.20241016"
 description = "Typing stubs for requests"
 optional = false
 python-versions = ">=3.8"
-groups = ["main"]
+groups = ["main", "dev"]
 markers = "python_version == \"3.11\" or python_version >= \"3.12\""
 files = [
     {file = "types-requests-2.32.0.20241016.tar.gz", hash = "sha256:0d9cad2f27515d0e3e3da7134a1b6f28fb97129d86b867f24d9c726452634d95"},
@@ -10917,6 +11028,35 @@ files = [
 [package.dependencies]
 urllib3 = ">=2"
 
+[[package]]
+name = "types-six"
+version = "1.17.0.20241205"
+description = "Typing stubs for six"
+optional = false
+python-versions = ">=3.8"
+groups = ["dev"]
+markers = "python_version == \"3.11\" or python_version >= \"3.12\""
+files = [
+    {file = "types_six-1.17.0.20241205-py3-none-any.whl", hash = "sha256:a4947c2bdcd9ab69d44466a533a15839ff48ddc27223615cb8145d73ab805bc2"},
+    {file = "types_six-1.17.0.20241205.tar.gz", hash = "sha256:1f662347a8f3b2bf30517d629d82f591420df29811794b0bf3804e14d716f6e0"},
+]
+
+[[package]]
+name = "types-tqdm"
+version = "4.67.0.20241221"
+description = "Typing stubs for tqdm"
+optional = false
+python-versions = ">=3.8"
+groups = ["dev"]
+markers = "python_version == \"3.11\" or python_version >= \"3.12\""
+files = [
+    {file = "types_tqdm-4.67.0.20241221-py3-none-any.whl", hash = "sha256:a1f1c9cda5c2d8482d2c73957a5398bfdedda10f6bc7b3b4e812d5c910486d29"},
+    {file = "types_tqdm-4.67.0.20241221.tar.gz", hash = "sha256:e56046631056922385abe89aeb18af5611f471eadd7918a0ad7f34d84cd4c8cc"},
+]
+
+[package.dependencies]
+types-requests = "*"
+
 [[package]]
 name = "typing-extensions"
 version = "4.12.2"
@@ -11172,7 +11312,7 @@ version = "2.3.0"
 description = "HTTP library with thread-safe connection pooling, file post, and more."
 optional = false
 python-versions = ">=3.9"
-groups = ["main", "storage", "tools", "vdb"]
+groups = ["main", "dev", "storage", "tools", "vdb"]
 markers = "python_version == \"3.11\" or python_version >= \"3.12\""
 files = [
     {file = "urllib3-2.3.0-py3-none-any.whl", hash = "sha256:1cee9ad369867bfdbbb48b7dd50374c0967a0bb7710050facf0dd6911440e3df"},
@@ -11667,7 +11807,7 @@ version = "3.1.3"
 description = "The comprehensive WSGI web application library."
 optional = false
 python-versions = ">=3.9"
-groups = ["main", "tools"]
+groups = ["main", "dev", "tools"]
 markers = "python_version == \"3.11\" or python_version >= \"3.12\""
 files = [
     {file = "werkzeug-3.1.3-py3-none-any.whl", hash = "sha256:54b78bf3716d19a65be4fceccc0d1d7b89e608834989dfae50ea87564639213e"},
@@ -12248,4 +12388,4 @@ cffi = ["cffi (>=1.11)"]
 [metadata]
 lock-version = "2.1"
 python-versions = ">=3.11,<3.13"
-content-hash = "fdc2199389f0e4b6d81b4b7fe2c1d303b1995643fe802ad3a28b196e68c258ae"
+content-hash = "6243573a26b9aa03558eb2c176d2477a08b1033a17065e870e4be83af0af644d"

api/pyproject.toml

@@ -21,7 +21,7 @@ azure-ai-inference = "~1.0.0b3"
 azure-ai-ml = "~1.20.0"
 azure-identity = "1.16.1"
 beautifulsoup4 = "4.12.2"
-boto3 = "1.35.74"
+boto3 = "1.36.4"
 bs4 = "~0.0.1"
 cachetools = "~5.3.0"
 celery = "~5.4.0"
@@ -88,7 +88,6 @@ tencentcloud-sdk-python-hunyuan = "~3.0.1294"
 tiktoken = "~0.8.0"
 tokenizers = "~0.15.0"
 transformers = "~4.35.0"
-types-pytz = "~2024.2.0.20241003"
 unstructured = { version = "~0.16.1", extras = ["docx", "epub", "md", "msg", "ppt", "pptx"] }
 validators = "0.21.0"
 volcengine-python-sdk = {extras = ["ark"], version = "~1.0.98"}
@@ -183,6 +182,21 @@ pytest = "~8.3.2"
 pytest-benchmark = "~4.0.0"
 pytest-env = "~1.1.3"
 pytest-mock = "~3.14.0"
+types-beautifulsoup4 = "~4.12.0.20241020"
+types-flask-cors = "~5.0.0.20240902"
+types-flask-migrate = "~4.1.0.20250112"
+types-html5lib = "~1.1.11.20241018"
+types-openpyxl = "~3.1.5.20241225"
+types-protobuf = "~5.29.1.20241207"
+types-psutil = "~6.1.0.20241221"
+types-psycopg2 = "~2.9.21.20250121"
+types-python-dateutil = "~2.9.0.20241206"
+types-pytz = "~2024.2.0.20241221"
+types-pyyaml = "~6.0.12.20241230"
+types-regex = "~2024.11.6.20241221"
+types-requests = "~2.32.0.20241016"
+types-six = "~1.17.0.20241205"
+types-tqdm = "~4.67.0.20241221"
 
 ############################################################
 # [ Lint ] dependency group

api/services/auth/firecrawl/firecrawl.py

@@ -21,10 +21,12 @@ class FirecrawlAuth(ApiKeyAuthBase):
         headers = self._prepare_headers()
         options = {
             "url": "https://example.com",
-            "crawlerOptions": {"excludes": [], "includes": [], "limit": 1},
-            "pageOptions": {"onlyMainContent": True},
+            "includePaths": [],
+            "excludePaths": [],
+            "limit": 1,
+            "scrapeOptions": {"onlyMainContent": True},
         }
-        response = self._post_request(f"{self.base_url}/v0/crawl", options, headers)
+        response = self._post_request(f"{self.base_url}/v1/crawl", options, headers)
         if response.status_code == 200:
             return True
         else:

api/services/dataset_service.py

@@ -42,6 +42,7 @@ from models.source import DataSourceOauthBinding
 from services.entities.knowledge_entities.knowledge_entities import (
     ChildChunkUpdateArgs,
     KnowledgeConfig,
+    MetaDataConfig,
     RerankingModel,
     RetrievalModel,
     SegmentUpdateArgs,
@@ -894,6 +895,9 @@ class DocumentService:
                                 document.data_source_info = json.dumps(data_source_info)
                                 document.batch = batch
                                 document.indexing_status = "waiting"
+                                if knowledge_config.metadata:
+                                    document.doc_type = knowledge_config.metadata.doc_type
+                                    document.metadata = knowledge_config.metadata.doc_metadata
                                 db.session.add(document)
                                 documents.append(document)
                                 duplicate_document_ids.append(document.id)
@@ -910,6 +914,7 @@ class DocumentService:
                             account,
                             file_name,
                             batch,
+                            knowledge_config.metadata,
                         )
                         db.session.add(document)
                         db.session.flush()
@@ -965,6 +970,7 @@ class DocumentService:
                                     account,
                                     page.page_name,
                                     batch,
+                                    knowledge_config.metadata,
                                 )
                                 db.session.add(document)
                                 db.session.flush()
@@ -1005,6 +1011,7 @@ class DocumentService:
                             account,
                             document_name,
                             batch,
+                            knowledge_config.metadata,
                         )
                         db.session.add(document)
                         db.session.flush()
@@ -1042,6 +1049,7 @@ class DocumentService:
         account: Account,
         name: str,
         batch: str,
+        metadata: Optional[MetaDataConfig] = None,
     ):
         document = Document(
             tenant_id=dataset.tenant_id,
@@ -1057,6 +1065,9 @@ class DocumentService:
             doc_form=document_form,
             doc_language=document_language,
         )
+        if metadata is not None:
+            document.doc_metadata = metadata.doc_metadata
+            document.doc_type = metadata.doc_type
         return document
 
     @staticmethod
@@ -1169,6 +1180,10 @@ class DocumentService:
         # update document name
         if document_data.name:
             document.name = document_data.name
+        # update doc_type and doc_metadata if provided
+        if document_data.metadata is not None:
+            document.doc_metadata = document_data.metadata.doc_type
+            document.doc_type = document_data.metadata.doc_type
         # update document to be waiting
         document.indexing_status = "waiting"
         document.completed_at = None

api/services/entities/knowledge_entities/knowledge_entities.py

@@ -93,6 +93,11 @@ class RetrievalModel(BaseModel):
     score_threshold: Optional[float] = None
 
 
+class MetaDataConfig(BaseModel):
+    doc_type: str
+    doc_metadata: dict
+
+
 class KnowledgeConfig(BaseModel):
     original_document_id: Optional[str] = None
     duplicate: bool = True
@@ -105,6 +110,7 @@ class KnowledgeConfig(BaseModel):
     embedding_model: Optional[str] = None
     embedding_model_provider: Optional[str] = None
     name: Optional[str] = None
+    metadata: Optional[MetaDataConfig] = None
 
 
 class SegmentUpdateArgs(BaseModel):

api/services/website_service.py

@@ -38,30 +38,22 @@ class WebsiteService:
             only_main_content = options.get("only_main_content", False)
             if not crawl_sub_pages:
                 params = {
-                    "crawlerOptions": {
-                        "includes": [],
-                        "excludes": [],
-                        "generateImgAltText": True,
-                        "limit": 1,
-                        "returnOnlyUrls": False,
-                        "pageOptions": {"onlyMainContent": only_main_content, "includeHtml": False},
-                    }
+                    "includePaths": [],
+                    "excludePaths": [],
+                    "limit": 1,
+                    "scrapeOptions": {"onlyMainContent": only_main_content},
                 }
             else:
                 includes = options.get("includes").split(",") if options.get("includes") else []
                 excludes = options.get("excludes").split(",") if options.get("excludes") else []
                 params = {
-                    "crawlerOptions": {
-                        "includes": includes,
-                        "excludes": excludes,
-                        "generateImgAltText": True,
-                        "limit": options.get("limit", 1),
-                        "returnOnlyUrls": False,
-                        "pageOptions": {"onlyMainContent": only_main_content, "includeHtml": False},
-                    }
+                    "includePaths": includes,
+                    "excludePaths": excludes,
+                    "limit": options.get("limit", 1),
+                    "scrapeOptions": {"onlyMainContent": only_main_content},
                 }
                 if options.get("max_depth"):
-                    params["crawlerOptions"]["maxDepth"] = options.get("max_depth")
+                    params["maxDepth"] = options.get("max_depth")
             job_id = firecrawl_app.crawl_url(url, params)
             website_crawl_time_cache_key = f"website_crawl_{job_id}"
             time = str(datetime.datetime.now().timestamp())
@@ -228,7 +220,7 @@ class WebsiteService:
             # decrypt api_key
             api_key = encrypter.decrypt_token(tenant_id=tenant_id, token=credentials.get("config").get("api_key"))
             firecrawl_app = FirecrawlApp(api_key=api_key, base_url=credentials.get("config").get("base_url", None))
-            params = {"pageOptions": {"onlyMainContent": only_main_content, "includeHtml": False}}
+            params = {"onlyMainContent": only_main_content}
             result = firecrawl_app.scrape_url(url, params)
             return result
         else:

api/tests/integration_tests/model_runtime/huggingface_hub/test_llm.py

@@ -10,6 +10,7 @@ from core.model_runtime.model_providers.huggingface_hub.llm.llm import Huggingfa
 from tests.integration_tests.model_runtime.__mock.huggingface import setup_huggingface_mock
 
 
+@pytest.mark.skip
 @pytest.mark.parametrize("setup_huggingface_mock", [["none"]], indirect=True)
 def test_hosted_inference_api_validate_credentials(setup_huggingface_mock):
     model = HuggingfaceHubLargeLanguageModel()

api/tests/unit_tests/core/rag/extractor/firecrawl/test_firecrawl.py

@@ -10,19 +10,17 @@ def test_firecrawl_web_extractor_crawl_mode(mocker):
     base_url = "https://api.firecrawl.dev"
     firecrawl_app = FirecrawlApp(api_key=api_key, base_url=base_url)
     params = {
-        "crawlerOptions": {
-            "includes": [],
-            "excludes": [],
-            "generateImgAltText": True,
-            "maxDepth": 1,
-            "limit": 1,
-            "returnOnlyUrls": False,
-        }
+        "includePaths": [],
+        "excludePaths": [],
+        "maxDepth": 1,
+        "limit": 1,
     }
     mocked_firecrawl = {
-        "jobId": "test",
+        "id": "test",
     }
     mocker.patch("requests.post", return_value=_mock_response(mocked_firecrawl))
     job_id = firecrawl_app.crawl_url(url, params)
-    print(job_id)
+    print(f"job_id: {job_id}")
+
+    assert job_id is not None
     assert isinstance(job_id, str)

web/.eslintignore

@@ -1,7 +0,0 @@
-/**/node_modules/*
-node_modules/
-
-dist/
-build/
-out/
-.next/

web/.eslintrc.json

@@ -1,31 +0,0 @@
-{
-  "extends": [
-    "next",
-    "@antfu",
-    "plugin:storybook/recommended"
-  ],
-  "rules": {
-    "@typescript-eslint/consistent-type-definitions": [
-      "error",
-      "type"
-    ],
-    "@typescript-eslint/no-var-requires": "off",
-    "no-console": "off",
-    "indent": "off",
-    "@typescript-eslint/indent": [
-      "error",
-      2,
-      {
-        "SwitchCase": 1,
-        "flatTernaryExpressions": false,
-        "ignoredNodes": [
-          "PropertyDefinition[decorators]",
-          "TSUnionType",
-          "FunctionExpression[params]:has(Identifier[decorators])"
-        ]
-      }
-    ],
-    "react-hooks/exhaustive-deps": "warn",
-    "react/display-name": "warn"
-  }
-}

web/app/(commonLayout)/datasets/template/template.en.mdx

@@ -47,6 +47,44 @@ import { Row, Col, Properties, Property, Heading, SubProperty, PropertyInstructi
       <Property name='text' type='string' key='text'>
         Document content
       </Property>
+      <Property name='doc_type' type='string' key='doc_type'>
+        Type of document (optional):
+          - <code>book</code> Book
+          - <code>web_page</code> Web page
+          - <code>paper</code> Academic paper/article 
+          - <code>social_media_post</code> Social media post
+          - <code>wikipedia_entry</code> Wikipedia entry
+          - <code>personal_document</code> Personal document
+          - <code>business_document</code> Business document
+          - <code>im_chat_log</code> Chat log
+          - <code>synced_from_notion</code> Notion document
+          - <code>synced_from_github</code> GitHub document
+          - <code>others</code> Other document types
+      </Property>
+      <Property name='doc_metadata' type='object' key='doc_metadata'>
+        Document metadata (required if doc_type is provided). Fields vary by doc_type:
+          For <code>book</code>:
+          - <code>title</code> Book title 
+          - <code>language</code> Book language
+          - <code>author</code> Book author
+          - <code>publisher</code> Publisher name
+          - <code>publication_date</code> Publication date
+          - <code>isbn</code> ISBN number
+          - <code>category</code> Book category
+
+          For <code>web_page</code>:
+          - <code>title</code> Page title
+          - <code>url</code> Page URL
+          - <code>language</code> Page language
+          - <code>publish_date</code> Publish date
+          - <code>author/publisher</code> Author or publisher
+          - <code>topic/keywords</code> Topic or keywords
+          - <code>description</code> Page description
+
+          Please check [api/services/dataset_service.py](https://github.com/langgenius/dify/blob/main/api/services/dataset_service.py#L475) for more details on the fields required for each doc_type.
+
+          For doc_type "others", any valid JSON object is accepted
+      </Property>
       <Property name='indexing_technique' type='string' key='indexing_technique'>
         Index mode
           - <code>high_quality</code> High quality: embedding using embedding model, built as vector database index
@@ -195,6 +233,68 @@ import { Row, Col, Properties, Property, Heading, SubProperty, PropertyInstructi
           - <code>hierarchical_model</code> Parent-child mode
           - <code>qa_model</code> Q&A Mode: Generates Q&A pairs for segmented documents and then embeds the questions
 
+        - <code>doc_type</code> Type of document (optional)
+          - <code>book</code> Book
+            Document records a book or publication
+          - <code>web_page</code> Web page 
+            Document records web page content
+          - <code>paper</code> Academic paper/article
+            Document records academic paper or research article
+          - <code>social_media_post</code> Social media post
+            Content from social media posts
+          - <code>wikipedia_entry</code> Wikipedia entry
+            Content from Wikipedia entries
+          - <code>personal_document</code> Personal document
+            Documents related to personal content
+          - <code>business_document</code> Business document
+            Documents related to business content
+          - <code>im_chat_log</code> Chat log
+            Records of instant messaging chats
+          - <code>synced_from_notion</code> Notion document
+            Documents synchronized from Notion
+          - <code>synced_from_github</code> GitHub document
+            Documents synchronized from GitHub
+          - <code>others</code> Other document types
+            Other document types not listed above
+
+        - <code>doc_metadata</code> Document metadata (required if doc_type is provided)
+          Fields vary by doc_type:
+
+          For <code>book</code>:
+          - <code>title</code> Book title
+            Title of the book
+          - <code>language</code> Book language
+            Language of the book
+          - <code>author</code> Book author
+            Author of the book
+          - <code>publisher</code> Publisher name
+            Name of the publishing house
+          - <code>publication_date</code> Publication date
+            Date when the book was published
+          - <code>isbn</code> ISBN number
+            International Standard Book Number
+          - <code>category</code> Book category
+            Category or genre of the book
+
+          For <code>web_page</code>:
+          - <code>title</code> Page title
+            Title of the web page
+          - <code>url</code> Page URL
+            URL address of the web page
+          - <code>language</code> Page language
+            Language of the web page
+          - <code>publish_date</code> Publish date
+            Date when the web page was published
+          - <code>author/publisher</code> Author or publisher
+            Author or publisher of the web page
+          - <code>topic/keywords</code> Topic or keywords
+            Topics or keywords of the web page
+          - <code>description</code> Page description
+            Description of the web page content
+
+          Please check [api/services/dataset_service.py](https://github.com/langgenius/dify/blob/main/api/services/dataset_service.py#L475) for more details on the fields required for each doc_type.
+          For doc_type "others", any valid JSON object is accepted
+
         - <code>doc_language</code> In Q&A mode, specify the language of the document, for example: <code>English</code>, <code>Chinese</code>
 
         - <code>process_rule</code> Processing rules
@@ -307,6 +407,44 @@ import { Row, Col, Properties, Property, Heading, SubProperty, PropertyInstructi
       <Property name='description' type='string' key='description'>
         Knowledge description (optional)
       </Property>
+      <Property name='doc_type' type='string' key='doc_type'>
+        Type of document (optional):
+          - <code>book</code> Book
+          - <code>web_page</code> Web page
+          - <code>paper</code> Academic paper/article 
+          - <code>social_media_post</code> Social media post
+          - <code>wikipedia_entry</code> Wikipedia entry
+          - <code>personal_document</code> Personal document
+          - <code>business_document</code> Business document
+          - <code>im_chat_log</code> Chat log
+          - <code>synced_from_notion</code> Notion document
+          - <code>synced_from_github</code> GitHub document
+          - <code>others</code> Other document types
+      </Property>
+      <Property name='doc_metadata' type='object' key='doc_metadata'>
+        Document metadata (required if doc_type is provided). Fields vary by doc_type:
+          For <code>book</code>:
+          - <code>title</code> Book title 
+          - <code>language</code> Book language
+          - <code>author</code> Book author
+          - <code>publisher</code> Publisher name
+          - <code>publication_date</code> Publication date
+          - <code>isbn</code> ISBN number
+          - <code>category</code> Book category
+
+          For <code>web_page</code>:
+          - <code>title</code> Page title
+          - <code>url</code> Page URL
+          - <code>language</code> Page language
+          - <code>publish_date</code> Publish date
+          - <code>author/publisher</code> Author or publisher
+          - <code>topic/keywords</code> Topic or keywords
+          - <code>description</code> Page description
+
+          Please check [api/services/dataset_service.py](https://github.com/langgenius/dify/blob/main/api/services/dataset_service.py#L475) for more details on the fields required for each doc_type.
+
+          For doc_type "others", any valid JSON object is accepted
+      </Property>
       <Property name='indexing_technique' type='string' key='indexing_technique'>
         Index technique (optional)
           - <code>high_quality</code> High quality
@@ -624,6 +762,67 @@ import { Row, Col, Properties, Property, Heading, SubProperty, PropertyInstructi
               - <code>separator</code> Segmentation identifier. Currently, only one delimiter is allowed. The default is <code>***</code>
               - <code>max_tokens</code> The maximum length (tokens) must be validated to be shorter than the length of the parent chunk
               - <code>chunk_overlap</code> Define the overlap between adjacent chunks (optional)
+            - <code>doc_type</code> Type of document (optional)
+              - <code>book</code> Book
+                Document records a book or publication
+              - <code>web_page</code> Web page 
+                Document records web page content
+              - <code>paper</code> Academic paper/article
+                Document records academic paper or research article
+              - <code>social_media_post</code> Social media post
+                Content from social media posts
+              - <code>wikipedia_entry</code> Wikipedia entry
+                Content from Wikipedia entries
+              - <code>personal_document</code> Personal document
+                Documents related to personal content
+              - <code>business_document</code> Business document
+                Documents related to business content
+              - <code>im_chat_log</code> Chat log
+                Records of instant messaging chats
+              - <code>synced_from_notion</code> Notion document
+                Documents synchronized from Notion
+              - <code>synced_from_github</code> GitHub document
+                Documents synchronized from GitHub
+              - <code>others</code> Other document types
+                Other document types not listed above
+
+            - <code>doc_metadata</code> Document metadata (required if doc_type is provided)
+              Fields vary by doc_type:
+
+              For <code>book</code>:
+              - <code>title</code> Book title
+                Title of the book
+              - <code>language</code> Book language
+                Language of the book
+              - <code>author</code> Book author
+                Author of the book
+              - <code>publisher</code> Publisher name
+                Name of the publishing house
+              - <code>publication_date</code> Publication date
+                Date when the book was published
+              - <code>isbn</code> ISBN number
+                International Standard Book Number
+              - <code>category</code> Book category
+                Category or genre of the book
+
+              For <code>web_page</code>:
+              - <code>title</code> Page title
+                Title of the web page
+              - <code>url</code> Page URL
+                URL address of the web page
+              - <code>language</code> Page language
+                Language of the web page
+              - <code>publish_date</code> Publish date
+                Date when the web page was published
+              - <code>author/publisher</code> Author or publisher
+                Author or publisher of the web page
+              - <code>topic/keywords</code> Topic or keywords
+                Topics or keywords of the web page
+              - <code>description</code> Page description
+                Description of the web page content
+
+              Please check [api/services/dataset_service.py](https://github.com/langgenius/dify/blob/main/api/services/dataset_service.py#L475) for more details on the fields required for each doc_type.
+              For doc_type "others", any valid JSON object is accepted
       </Property>
     </Properties>
   </Col>

web/app/(commonLayout)/datasets/template/template.zh.mdx

@@ -47,6 +47,46 @@ import { Row, Col, Properties, Property, Heading, SubProperty, PropertyInstructi
       <Property name='text' type='string' key='text'>
         文档内容
       </Property>
+      <Property name='doc_type' type='string' key='doc_type'>
+        文档类型（选填）
+          - <code>book</code> 图书 Book
+          - <code>web_page</code> 网页 Web page
+          - <code>paper</code> 学术论文/文章 Academic paper/article 
+          - <code>social_media_post</code> 社交媒体帖子 Social media post
+          - <code>wikipedia_entry</code> 维基百科条目 Wikipedia entry
+          - <code>personal_document</code> 个人文档 Personal document
+          - <code>business_document</code> 商业文档 Business document
+          - <code>im_chat_log</code> 即时通讯记录 Chat log
+          - <code>synced_from_notion</code> Notion同步文档 Notion document
+          - <code>synced_from_github</code> GitHub同步文档 GitHub document
+          - <code>others</code> 其他文档类型 Other document types
+      </Property>
+      <Property name='doc_metadata' type='object' key='doc_metadata'>
+      
+        文档元数据（如提供文档类型则必填）。字段因文档类型而异：
+          
+          针对图书 For <code>book</code>:
+          - <code>title</code> 书名 Book title 
+          - <code>language</code> 图书语言 Book language
+          - <code>author</code> 作者 Book author
+          - <code>publisher</code> 出版社 Publisher name
+          - <code>publication_date</code> 出版日期 Publication date
+          - <code>isbn</code> ISBN号码 ISBN number
+          - <code>category</code> 图书分类 Book category
+
+          针对网页 For <code>web_page</code>:
+          - <code>title</code> 页面标题 Page title
+          - <code>url</code> 页面网址 Page URL
+          - <code>language</code> 页面语言 Page language
+          - <code>publish_date</code> 发布日期 Publish date
+          - <code>author/publisher</code> 作者/发布者 Author or publisher
+          - <code>topic/keywords</code> 主题/关键词 Topic or keywords
+          - <code>description</code> 页面描述 Page description
+
+          请查看 [api/services/dataset_service.py](https://github.com/langgenius/dify/blob/main/api/services/dataset_service.py#L475) 了解各文档类型所需字段的详细信息。
+
+          针对"其他"类型文档，接受任何有效的JSON对象
+      </Property>
       <Property name='indexing_technique' type='string' key='indexing_technique'>
         索引方式
           - <code>high_quality</code> 高质量：使用  embedding 模型进行嵌入，构建为向量数据库索引
@@ -194,6 +234,68 @@ import { Row, Col, Properties, Property, Heading, SubProperty, PropertyInstructi
           - <code>text_model</code> text 文档直接 embedding，经济模式默认为该模式
           - <code>hierarchical_model</code> parent-child 模式
           - <code>qa_model</code> Q&A 模式：为分片文档生成 Q&A 对，然后对问题进行 embedding
+        - <code>doc_type</code> 文档类型（选填）Type of document (optional)
+          - <code>book</code> 图书
+            文档记录一本书籍或出版物
+          - <code>web_page</code> 网页
+            网页内容的文档记录
+          - <code>paper</code> 学术论文/文章
+            学术论文或研究文章的记录
+          - <code>social_media_post</code> 社交媒体帖子
+            社交媒体上的帖子内容
+          - <code>wikipedia_entry</code> 维基百科条目
+            维基百科的词条内容
+          - <code>personal_document</code> 个人文档
+            个人相关的文档记录
+          - <code>business_document</code> 商业文档
+            商业相关的文档记录
+          - <code>im_chat_log</code> 即时通讯记录
+            即时通讯的聊天记录
+          - <code>synced_from_notion</code> Notion同步文档
+            从Notion同步的文档内容
+          - <code>synced_from_github</code> GitHub同步文档
+            从GitHub同步的文档内容
+          - <code>others</code> 其他文档类型
+            其他未列出的文档类型
+
+        - <code>doc_metadata</code> 文档元数据（如提供文档类型则必填
+          字段因文档类型而异
+
+          针对图书类型 For <code>book</code>:
+          - <code>title</code> 书名
+            书籍的标题
+          - <code>language</code> 图书语言
+            书籍的语言
+          - <code>author</code> 作者
+            书籍的作者
+          - <code>publisher</code> 出版社
+            出版社的名称
+          - <code>publication_date</code> 出版日期
+            书籍的出版日期
+          - <code>isbn</code> ISBN号码
+            书籍的ISBN编号
+          - <code>category</code> 图书分类
+            书籍的分类类别
+
+          针对网页类型 For <code>web_page</code>:
+          - <code>title</code> 页面标题
+            网页的标题
+          - <code>url</code> 页面网址
+            网页的URL地址
+          - <code>language</code> 页面语言
+            网页的语言
+          - <code>publish_date</code> 发布日期
+            网页的发布日期
+          - <code>author/publisher</code> 作者/发布者
+            网页的作者或发布者
+          - <code>topic/keywords</code> 主题/关键词
+            网页的主题或关键词
+          - <code>description</code> 页面描述
+            网页的描述信息
+
+          请查看 [api/services/dataset_service.py](https://github.com/langgenius/dify/blob/main/api/services/dataset_service.py#L475) 了解各文档类型所需字段的详细信息。
+
+          针对"其他"类型文档，接受任何有效的JSON对象
 
         - <code>doc_language</code> 在 Q&A 模式下，指定文档的语言，例如：<code>English</code>、<code>Chinese</code>
 
@@ -504,6 +606,46 @@ import { Row, Col, Properties, Property, Heading, SubProperty, PropertyInstructi
       <Property name='text' type='string' key='text'>
         文档内容（选填）
       </Property>
+      <Property name='doc_type' type='string' key='doc_type'>
+        文档类型（选填）
+          - <code>book</code> 图书 Book
+          - <code>web_page</code> 网页 Web page
+          - <code>paper</code> 学术论文/文章 Academic paper/article 
+          - <code>social_media_post</code> 社交媒体帖子 Social media post
+          - <code>wikipedia_entry</code> 维基百科条目 Wikipedia entry
+          - <code>personal_document</code> 个人文档 Personal document
+          - <code>business_document</code> 商业文档 Business document
+          - <code>im_chat_log</code> 即时通讯记录 Chat log
+          - <code>synced_from_notion</code> Notion同步文档 Notion document
+          - <code>synced_from_github</code> GitHub同步文档 GitHub document
+          - <code>others</code> 其他文档类型 Other document types
+      </Property>
+      <Property name='doc_metadata' type='object' key='doc_metadata'>
+      
+        文档元数据（如提供文档类型则必填）。字段因文档类型而异：
+          
+          针对图书 For <code>book</code>:
+          - <code>title</code> 书名 Book title 
+          - <code>language</code> 图书语言 Book language
+          - <code>author</code> 作者 Book author
+          - <code>publisher</code> 出版社 Publisher name
+          - <code>publication_date</code> 出版日期 Publication date
+          - <code>isbn</code> ISBN号码 ISBN number
+          - <code>category</code> 图书分类 Book category
+
+          针对网页 For <code>web_page</code>:
+          - <code>title</code> 页面标题 Page title
+          - <code>url</code> 页面网址 Page URL
+          - <code>language</code> 页面语言 Page language
+          - <code>publish_date</code> 发布日期 Publish date
+          - <code>author/publisher</code> 作者/发布者 Author or publisher
+          - <code>topic/keywords</code> 主题/关键词 Topic or keywords
+          - <code>description</code> 页面描述 Page description
+
+          请查看 [api/services/dataset_service.py](https://github.com/langgenius/dify/blob/main/api/services/dataset_service.py#L475) 了解各文档类型所需字段的详细信息。
+
+          针对"其他"类型文档，接受任何有效的JSON对象
+      </Property>
       <Property name='process_rule' type='object' key='process_rule'>
         处理规则（选填）
           - <code>mode</code> (string) 清洗、分段模式 ，automatic 自动 / custom 自定义
@@ -624,6 +766,68 @@ import { Row, Col, Properties, Property, Heading, SubProperty, PropertyInstructi
               - <code>separator</code> 分段标识符，目前仅允许设置一个分隔符。默认为 <code>***</code>
               - <code>max_tokens</code> 最大长度 (token) 需要校验小于父级的长度
               - <code>chunk_overlap</code> 分段重叠指的是在对数据进行分段时，段与段之间存在一定的重叠部分（选填）
+            - <code>doc_type</code> 文档类型（选填）Type of document (optional)
+              - <code>book</code> 图书
+                文档记录一本书籍或出版物
+              - <code>web_page</code> 网页
+                网页内容的文档记录
+              - <code>paper</code> 学术论文/文章
+                学术论文或研究文章的记录
+              - <code>social_media_post</code> 社交媒体帖子
+                社交媒体上的帖子内容
+              - <code>wikipedia_entry</code> 维基百科条目
+                维基百科的词条内容
+              - <code>personal_document</code> 个人文档
+                个人相关的文档记录
+              - <code>business_document</code> 商业文档
+                商业相关的文档记录
+              - <code>im_chat_log</code> 即时通讯记录
+                即时通讯的聊天记录
+              - <code>synced_from_notion</code> Notion同步文档
+                从Notion同步的文档内容
+              - <code>synced_from_github</code> GitHub同步文档
+                从GitHub同步的文档内容
+              - <code>others</code> 其他文档类型
+                其他未列出的文档类型
+
+            - <code>doc_metadata</code> 文档元数据（如提供文档类型则必填
+              字段因文档类型而异
+
+              针对图书类型 For <code>book</code>:
+              - <code>title</code> 书名
+                书籍的标题
+              - <code>language</code> 图书语言
+                书籍的语言
+              - <code>author</code> 作者
+                书籍的作者
+              - <code>publisher</code> 出版社
+                出版社的名称
+              - <code>publication_date</code> 出版日期
+                书籍的出版日期
+              - <code>isbn</code> ISBN号码
+                书籍的ISBN编号
+              - <code>category</code> 图书分类
+                书籍的分类类别
+
+              针对网页类型 For <code>web_page</code>:
+              - <code>title</code> 页面标题
+                网页的标题
+              - <code>url</code> 页面网址
+                网页的URL地址
+              - <code>language</code> 页面语言
+                网页的语言
+              - <code>publish_date</code> 发布日期
+                网页的发布日期
+              - <code>author/publisher</code> 作者/发布者
+                网页的作者或发布者
+              - <code>topic/keywords</code> 主题/关键词
+                网页的主题或关键词
+              - <code>description</code> 页面描述
+                网页的描述信息
+
+              请查看 [api/services/dataset_service.py](https://github.com/langgenius/dify/blob/main/api/services/dataset_service.py#L475) 了解各文档类型所需字段的详细信息。
+
+              针对"其他"类型文档，接受任何有效的JSON对象
       </Property>
     </Properties>
   </Col>

web/app/components/app/configuration/debug/debug-with-multiple-model/chat-item.tsx

@@ -67,7 +67,6 @@ const ChatItem: FC<ChatItemProps> = ({
   }, [modelConfig.configs.prompt_variables])
   const {
     chatList,
-    chatListRef,
     isResponding,
     handleSend,
     suggestedQuestions,
@@ -102,7 +101,7 @@ const ChatItem: FC<ChatItemProps> = ({
       query: message,
       inputs,
       model_config: configData,
-      parent_message_id: getLastAnswer(chatListRef.current)?.id || null,
+      parent_message_id: getLastAnswer(chatList)?.id || null,
     }
 
     if ((config.file_upload as any).enabled && files?.length && supportVision)
@@ -116,7 +115,7 @@ const ChatItem: FC<ChatItemProps> = ({
         onGetSuggestedQuestions: (responseItemId, getAbortController) => fetchSuggestedQuestions(appId, responseItemId, getAbortController),
       },
     )
-  }, [appId, config, handleSend, inputs, modelAndParameter, textGenerationModelList, chatListRef])
+  }, [appId, chatList, config, handleSend, inputs, modelAndParameter.model, modelAndParameter.parameters, modelAndParameter.provider, textGenerationModelList])
 
   const { eventEmitter } = useEventEmitterContextContext()
   eventEmitter?.useSubscription((v: any) => {

web/app/components/app/configuration/debug/debug-with-single-model/index.tsx

@@ -12,7 +12,7 @@ import {
 import Chat from '@/app/components/base/chat/chat'
 import { useChat } from '@/app/components/base/chat/chat/hooks'
 import { useDebugConfigurationContext } from '@/context/debug-configuration'
-import type { ChatConfig, ChatItem, OnSend } from '@/app/components/base/chat/types'
+import type { ChatConfig, ChatItem, ChatItemInTree, OnSend } from '@/app/components/base/chat/types'
 import { useProviderContext } from '@/context/provider-context'
 import {
   fetchConversationMessages,
@@ -24,7 +24,7 @@ import { useAppContext } from '@/context/app-context'
 import { ModelFeatureEnum } from '@/app/components/header/account-setting/model-provider-page/declarations'
 import { useStore as useAppStore } from '@/app/components/app/store'
 import { useFeatures } from '@/app/components/base/features/hooks'
-import { getLastAnswer } from '@/app/components/base/chat/utils'
+import { getLastAnswer, isValidGeneratedAnswer } from '@/app/components/base/chat/utils'
 import type { InputForm } from '@/app/components/base/chat/chat/type'
 
 type DebugWithSingleModelProps = {
@@ -68,12 +68,11 @@ const DebugWithSingleModel = forwardRef<DebugWithSingleModelRefType, DebugWithSi
   }, [modelConfig.configs.prompt_variables])
   const {
     chatList,
-    chatListRef,
+    setTargetMessageId,
     isResponding,
     handleSend,
     suggestedQuestions,
     handleStop,
-    handleUpdateChatList,
     handleRestart,
     handleAnnotationAdded,
     handleAnnotationEdited,
@@ -89,7 +88,7 @@ const DebugWithSingleModel = forwardRef<DebugWithSingleModelRefType, DebugWithSi
   )
   useFormattingChangedSubscription(chatList)
 
-  const doSend: OnSend = useCallback((message, files, last_answer) => {
+  const doSend: OnSend = useCallback((message, files, isRegenerate = false, parentAnswer: ChatItem | null = null) => {
     if (checkCanSend && !checkCanSend())
       return
     const currentProvider = textGenerationModelList.find(item => item.provider === modelConfig.provider)
@@ -110,7 +109,7 @@ const DebugWithSingleModel = forwardRef<DebugWithSingleModelRefType, DebugWithSi
       query: message,
       inputs,
       model_config: configData,
-      parent_message_id: last_answer?.id || getLastAnswer(chatListRef.current)?.id || null,
+      parent_message_id: (isRegenerate ? parentAnswer?.id : getLastAnswer(chatList)?.id) || null,
     }
 
     if ((config.file_upload as any)?.enabled && files?.length && supportVision)
@@ -124,23 +123,13 @@ const DebugWithSingleModel = forwardRef<DebugWithSingleModelRefType, DebugWithSi
         onGetSuggestedQuestions: (responseItemId, getAbortController) => fetchSuggestedQuestions(appId, responseItemId, getAbortController),
       },
     )
-  }, [chatListRef, appId, checkCanSend, completionParams, config, handleSend, inputs, modelConfig, textGenerationModelList])
+  }, [appId, chatList, checkCanSend, completionParams, config, handleSend, inputs, modelConfig.mode, modelConfig.model_id, modelConfig.provider, textGenerationModelList])
 
-  const doRegenerate = useCallback((chatItem: ChatItem) => {
-    const index = chatList.findIndex(item => item.id === chatItem.id)
-    if (index === -1)
-      return
-
-    const prevMessages = chatList.slice(0, index)
-    const question = prevMessages.pop()
-    const lastAnswer = getLastAnswer(prevMessages)
-
-    if (!question)
-      return
-
-    handleUpdateChatList(prevMessages)
-    doSend(question.content, question.message_files, lastAnswer)
-  }, [chatList, handleUpdateChatList, doSend])
+  const doRegenerate = useCallback((chatItem: ChatItemInTree) => {
+    const question = chatList.find(item => item.id === chatItem.parentMessageId)!
+    const parentAnswer = chatList.find(item => item.id === question.parentMessageId)
+    doSend(question.content, question.message_files, true, isValidGeneratedAnswer(parentAnswer) ? parentAnswer : null)
+  }, [chatList, doSend])
 
   const allToolIcons = useMemo(() => {
     const icons: Record<string, any> = {}
@@ -173,6 +162,7 @@ const DebugWithSingleModel = forwardRef<DebugWithSingleModelRefType, DebugWithSi
       inputs={inputs}
       inputsForm={inputsForm}
       onRegenerate={doRegenerate}
+      switchSibling={siblingMessageId => setTargetMessageId(siblingMessageId)}
       onStopResponding={handleStop}
       showPromptLog
       questionIcon={<Avatar avatar={userProfile.avatar_url} name={userProfile.name} size={40} />}

web/app/components/base/chat/chat-with-history/chat-wrapper.tsx

@@ -3,10 +3,11 @@ import Chat from '../chat'
 import type {
   ChatConfig,
   ChatItem,
+  ChatItemInTree,
   OnSend,
 } from '../types'
 import { useChat } from '../chat/hooks'
-import { getLastAnswer } from '../utils'
+import { getLastAnswer, isValidGeneratedAnswer } from '../utils'
 import { useChatWithHistoryContext } from './context'
 import Header from './header'
 import ConfigPanel from './config-panel'
@@ -20,7 +21,7 @@ import AnswerIcon from '@/app/components/base/answer-icon'
 const ChatWrapper = () => {
   const {
     appParams,
-    appPrevChatList,
+    appPrevChatTree,
     currentConversationId,
     currentConversationItem,
     inputsForms,
@@ -50,8 +51,7 @@ const ChatWrapper = () => {
   }, [appParams, currentConversationItem?.introduction, currentConversationId])
   const {
     chatList,
-    chatListRef,
-    handleUpdateChatList,
+    setTargetMessageId,
     handleSend,
     handleStop,
     isResponding,
@@ -62,7 +62,7 @@ const ChatWrapper = () => {
       inputs: (currentConversationId ? currentConversationItem?.inputs : newConversationInputs) as any,
       inputsForm: inputsForms,
     },
-    appPrevChatList,
+    appPrevChatTree,
     taskId => stopChatMessageResponding('', taskId, isInstalledApp, appId),
   )
 
@@ -72,13 +72,13 @@ const ChatWrapper = () => {
   // eslint-disable-next-line react-hooks/exhaustive-deps
   }, [])
 
-  const doSend: OnSend = useCallback((message, files, last_answer) => {
+  const doSend: OnSend = useCallback((message, files, isRegenerate = false, parentAnswer: ChatItem | null = null) => {
     const data: any = {
       query: message,
       files,
       inputs: currentConversationId ? currentConversationItem?.inputs : newConversationInputs,
       conversation_id: currentConversationId,
-      parent_message_id: last_answer?.id || getLastAnswer(chatListRef.current)?.id || null,
+      parent_message_id: (isRegenerate ? parentAnswer?.id : getLastAnswer(chatList)?.id) || null,
     }
 
     handleSend(
@@ -91,31 +91,21 @@ const ChatWrapper = () => {
       },
     )
   }, [
-    chatListRef,
+    chatList,
+    handleNewConversationCompleted,
+    handleSend,
     currentConversationId,
     currentConversationItem,
-    handleSend,
     newConversationInputs,
-    handleNewConversationCompleted,
     isInstalledApp,
     appId,
   ])
 
-  const doRegenerate = useCallback((chatItem: ChatItem) => {
-    const index = chatList.findIndex(item => item.id === chatItem.id)
-    if (index === -1)
-      return
-
-    const prevMessages = chatList.slice(0, index)
-    const question = prevMessages.pop()
-    const lastAnswer = getLastAnswer(prevMessages)
-
-    if (!question)
-      return
-
-    handleUpdateChatList(prevMessages)
-    doSend(question.content, question.message_files, lastAnswer)
-  }, [chatList, handleUpdateChatList, doSend])
+  const doRegenerate = useCallback((chatItem: ChatItemInTree) => {
+    const question = chatList.find(item => item.id === chatItem.parentMessageId)!
+    const parentAnswer = chatList.find(item => item.id === question.parentMessageId)
+    doSend(question.content, question.message_files, true, isValidGeneratedAnswer(parentAnswer) ? parentAnswer : null)
+  }, [chatList, doSend])
 
   const chatNode = useMemo(() => {
     if (inputsForms.length) {
@@ -187,6 +177,7 @@ const ChatWrapper = () => {
         answerIcon={answerIcon}
         hideProcessDetail
         themeBuilder={themeBuilder}
+        switchSibling={siblingMessageId => setTargetMessageId(siblingMessageId)}
       />
     </div>
   )

web/app/components/base/chat/chat-with-history/context.tsx

@@ -5,7 +5,7 @@ import { createContext, useContext } from 'use-context-selector'
 import type {
   Callback,
   ChatConfig,
-  ChatItem,
+  ChatItemInTree,
   Feedback,
 } from '../types'
 import type { ThemeBuilder } from '../embedded-chatbot/theme/theme-context'
@@ -25,7 +25,7 @@ export type ChatWithHistoryContextValue = {
   appChatListDataLoading?: boolean
   currentConversationId: string
   currentConversationItem?: ConversationItem
-  appPrevChatList: ChatItem[]
+  appPrevChatTree: ChatItemInTree[]
   pinnedConversationList: AppConversationData['data']
   conversationList: AppConversationData['data']
   showConfigPanelBeforeChat: boolean
@@ -53,7 +53,7 @@ export type ChatWithHistoryContextValue = {
 
 export const ChatWithHistoryContext = createContext<ChatWithHistoryContextValue>({
   currentConversationId: '',
-  appPrevChatList: [],
+  appPrevChatTree: [],
   pinnedConversationList: [],
   conversationList: [],
   showConfigPanelBeforeChat: false,

web/app/components/base/chat/chat-with-history/hooks.tsx

@@ -12,10 +12,13 @@ import produce from 'immer'
 import type {
   Callback,
   ChatConfig,
+  ChatItem,
   Feedback,
 } from '../types'
 import { CONVERSATION_ID_INFO } from '../constants'
-import { getPrevChatList } from '../utils'
+import { buildChatItemTree } from '../utils'
+import { addFileInfos, sortAgentSorts } from '../../../tools/utils'
+import { getProcessedFilesFromResponse } from '@/app/components/base/file-uploader/utils'
 import {
   delConversation,
   fetchAppInfo,
@@ -40,6 +43,32 @@ import { useAppFavicon } from '@/hooks/use-app-favicon'
 import { InputVarType } from '@/app/components/workflow/types'
 import { TransferMethod } from '@/types/app'
 
+function getFormattedChatList(messages: any[]) {
+  const newChatList: ChatItem[] = []
+  messages.forEach((item) => {
+    const questionFiles = item.message_files?.filter((file: any) => file.belongs_to === 'user') || []
+    newChatList.push({
+      id: `question-${item.id}`,
+      content: item.query,
+      isAnswer: false,
+      message_files: getProcessedFilesFromResponse(questionFiles.map((item: any) => ({ ...item, related_id: item.id }))),
+      parentMessageId: item.parent_message_id || undefined,
+    })
+    const answerFiles = item.message_files?.filter((file: any) => file.belongs_to === 'assistant') || []
+    newChatList.push({
+      id: item.id,
+      content: item.answer,
+      agent_thoughts: addFileInfos(item.agent_thoughts ? sortAgentSorts(item.agent_thoughts) : item.agent_thoughts, item.message_files),
+      feedback: item.feedback,
+      isAnswer: true,
+      citation: item.retriever_resources,
+      message_files: getProcessedFilesFromResponse(answerFiles.map((item: any) => ({ ...item, related_id: item.id }))),
+      parentMessageId: `question-${item.id}`,
+    })
+  })
+  return newChatList
+}
+
 export const useChatWithHistory = (installedAppInfo?: InstalledApp) => {
   const isInstalledApp = useMemo(() => !!installedAppInfo, [installedAppInfo])
   const { data: appInfo, isLoading: appInfoLoading, error: appInfoError } = useSWR(installedAppInfo ? null : 'appInfo', fetchAppInfo)
@@ -109,9 +138,9 @@ export const useChatWithHistory = (installedAppInfo?: InstalledApp) => {
   const { data: appConversationData, isLoading: appConversationDataLoading, mutate: mutateAppConversationData } = useSWR(['appConversationData', isInstalledApp, appId, false], () => fetchConversations(isInstalledApp, appId, undefined, false, 100))
   const { data: appChatListData, isLoading: appChatListDataLoading } = useSWR(chatShouldReloadKey ? ['appChatList', chatShouldReloadKey, isInstalledApp, appId] : null, () => fetchChatList(chatShouldReloadKey, isInstalledApp, appId))
 
-  const appPrevChatList = useMemo(
+  const appPrevChatTree = useMemo(
     () => (currentConversationId && appChatListData?.data.length)
-      ? getPrevChatList(appChatListData.data)
+      ? buildChatItemTree(getFormattedChatList(appChatListData.data))
       : [],
     [appChatListData, currentConversationId],
   )
@@ -403,7 +432,7 @@ export const useChatWithHistory = (installedAppInfo?: InstalledApp) => {
     appConversationDataLoading,
     appChatListData,
     appChatListDataLoading,
-    appPrevChatList,
+    appPrevChatTree,
     pinnedConversationList,
     conversationList,
     showConfigPanelBeforeChat,

web/app/components/base/chat/chat-with-history/index.tsx

@@ -30,7 +30,7 @@ const ChatWithHistory: FC<ChatWithHistoryProps> = ({
     appInfoError,
     appData,
     appInfoLoading,
-    appPrevChatList,
+    appPrevChatTree,
     showConfigPanelBeforeChat,
     appChatListDataLoading,
     chatShouldReloadKey,
@@ -38,7 +38,7 @@ const ChatWithHistory: FC<ChatWithHistoryProps> = ({
     themeBuilder,
   } = useChatWithHistoryContext()
 
-  const chatReady = (!showConfigPanelBeforeChat || !!appPrevChatList.length)
+  const chatReady = (!showConfigPanelBeforeChat || !!appPrevChatTree.length)
   const customConfig = appData?.custom_config
   const site = appData?.site
 
@@ -76,9 +76,9 @@ const ChatWithHistory: FC<ChatWithHistoryProps> = ({
           <HeaderInMobile />
         )
       }
-      <div className={`grow overflow-hidden ${showConfigPanelBeforeChat && !appPrevChatList.length && 'flex items-center justify-center'}`}>
+      <div className={`grow overflow-hidden ${showConfigPanelBeforeChat && !appPrevChatTree.length && 'flex items-center justify-center'}`}>
         {
-          showConfigPanelBeforeChat && !appChatListDataLoading && !appPrevChatList.length && (
+          showConfigPanelBeforeChat && !appChatListDataLoading && !appPrevChatTree.length && (
             <div className={`flex w-full items-center justify-center h-full ${isMobile && 'px-4'}`}>
               <ConfigPanel />
             </div>
@@ -120,7 +120,7 @@ const ChatWithHistoryWrap: FC<ChatWithHistoryWrapProps> = ({
     appChatListDataLoading,
     currentConversationId,
     currentConversationItem,
-    appPrevChatList,
+    appPrevChatTree,
     pinnedConversationList,
     conversationList,
     showConfigPanelBeforeChat,
@@ -154,7 +154,7 @@ const ChatWithHistoryWrap: FC<ChatWithHistoryWrapProps> = ({
       appChatListDataLoading,
       currentConversationId,
       currentConversationItem,
-      appPrevChatList,
+      appPrevChatTree,
       pinnedConversationList,
       conversationList,
       showConfigPanelBeforeChat,

web/app/components/base/chat/chat/answer/index.tsx

@@ -209,19 +209,19 @@ const Answer: FC<AnswerProps> = ({
             }
             {item.siblingCount && item.siblingCount > 1 && item.siblingIndex !== undefined && <div className="pt-3.5 flex justify-center items-center text-sm">
               <button
-                className={`${item.prevSibling ? 'opacity-100' : 'opacity-65'}`}
+                className={`${item.prevSibling ? 'opacity-100' : 'opacity-30'}`}
                 disabled={!item.prevSibling}
                 onClick={() => item.prevSibling && switchSibling?.(item.prevSibling)}
               >
-                <ChevronRight className="w-[14px] h-[14px] rotate-180 text-text-tertiary" />
+                <ChevronRight className="w-[14px] h-[14px] rotate-180 text-text-primary" />
               </button>
-              <span className="px-2 text-xs text-text-quaternary">{item.siblingIndex + 1} / {item.siblingCount}</span>
+              <span className="px-2 text-xs text-text-primary">{item.siblingIndex + 1} / {item.siblingCount}</span>
               <button
-                className={`${item.nextSibling ? 'opacity-100' : 'opacity-65'}`}
+                className={`${item.nextSibling ? 'opacity-100' : 'opacity-30'}`}
                 disabled={!item.nextSibling}
                 onClick={() => item.nextSibling && switchSibling?.(item.nextSibling)}
               >
-                <ChevronRight className="w-[14px] h-[14px] text-text-tertiary" />
+                <ChevronRight className="w-[14px] h-[14px] text-text-primary" />
               </button>
             </div>}
           </div>

web/app/components/base/chat/chat/hooks.ts

@@ -1,6 +1,7 @@
 import {
   useCallback,
   useEffect,
+  useMemo,
   useRef,
   useState,
 } from 'react'
@@ -12,8 +13,10 @@ import { v4 as uuidV4 } from 'uuid'
 import type {
   ChatConfig,
   ChatItem,
+  ChatItemInTree,
   Inputs,
 } from '../types'
+import { getThreadMessages } from '../utils'
 import type { InputForm } from './type'
 import {
   getProcessedInputs,
@@ -46,7 +49,7 @@ export const useChat = (
     inputs: Inputs
     inputsForm: InputForm[]
   },
-  prevChatList?: ChatItem[],
+  prevChatTree?: ChatItemInTree[],
   stopChat?: (taskId: string) => void,
 ) => {
   const { t } = useTranslation()
@@ -56,14 +59,48 @@ export const useChat = (
   const hasStopResponded = useRef(false)
   const [isResponding, setIsResponding] = useState(false)
   const isRespondingRef = useRef(false)
-  const [chatList, setChatList] = useState<ChatItem[]>(prevChatList || [])
-  const chatListRef = useRef<ChatItem[]>(prevChatList || [])
   const taskIdRef = useRef('')
   const [suggestedQuestions, setSuggestQuestions] = useState<string[]>([])
   const conversationMessagesAbortControllerRef = useRef<AbortController | null>(null)
   const suggestedQuestionsAbortControllerRef = useRef<AbortController | null>(null)
   const params = useParams()
   const pathname = usePathname()
+
+  const [chatTree, setChatTree] = useState<ChatItemInTree[]>(prevChatTree || [])
+  const chatTreeRef = useRef<ChatItemInTree[]>(chatTree)
+  const [targetMessageId, setTargetMessageId] = useState<string>()
+  const threadMessages = useMemo(() => getThreadMessages(chatTree, targetMessageId), [chatTree, targetMessageId])
+
+  const getIntroduction = useCallback((str: string) => {
+    return processOpeningStatement(str, formSettings?.inputs || {}, formSettings?.inputsForm || [])
+  }, [formSettings?.inputs, formSettings?.inputsForm])
+
+  /** Final chat list that will be rendered */
+  const chatList = useMemo(() => {
+    const ret = [...threadMessages]
+    if (config?.opening_statement) {
+      const index = threadMessages.findIndex(item => item.isOpeningStatement)
+
+      if (index > -1) {
+        ret[index] = {
+          ...ret[index],
+          content: getIntroduction(config.opening_statement),
+          suggestedQuestions: config.suggested_questions,
+        }
+      }
+      else {
+        ret.unshift({
+          id: `${Date.now()}`,
+          content: getIntroduction(config.opening_statement),
+          isAnswer: true,
+          isOpeningStatement: true,
+          suggestedQuestions: config.suggested_questions,
+        })
+      }
+    }
+    return ret
+  }, [threadMessages, config?.opening_statement, getIntroduction, config?.suggested_questions])
+
   useEffect(() => {
     setAutoFreeze(false)
     return () => {
@@ -71,43 +108,50 @@ export const useChat = (
     }
   }, [])
 
-  const handleUpdateChatList = useCallback((newChatList: ChatItem[]) => {
-    setChatList(newChatList)
-    chatListRef.current = newChatList
+  /** Find the target node by bfs and then operate on it */
+  const produceChatTreeNode = useCallback((targetId: string, operation: (node: ChatItemInTree) => void) => {
+    return produce(chatTreeRef.current, (draft) => {
+      const queue: ChatItemInTree[] = [...draft]
+      while (queue.length > 0) {
+        const current = queue.shift()!
+        if (current.id === targetId) {
+          operation(current)
+          break
+        }
+        if (current.children)
+          queue.push(...current.children)
+      }
+    })
   }, [])
+
+  type UpdateChatTreeNode = {
+    (id: string, fields: Partial<ChatItemInTree>): void
+    (id: string, update: (node: ChatItemInTree) => void): void
+  }
+
+  const updateChatTreeNode: UpdateChatTreeNode = useCallback((
+    id: string,
+    fieldsOrUpdate: Partial<ChatItemInTree> | ((node: ChatItemInTree) => void),
+  ) => {
+    const nextState = produceChatTreeNode(id, (node) => {
+      if (typeof fieldsOrUpdate === 'function') {
+        fieldsOrUpdate(node)
+      }
+      else {
+        Object.keys(fieldsOrUpdate).forEach((key) => {
+          (node as any)[key] = (fieldsOrUpdate as any)[key]
+        })
+      }
+    })
+    setChatTree(nextState)
+    chatTreeRef.current = nextState
+  }, [produceChatTreeNode])
+
   const handleResponding = useCallback((isResponding: boolean) => {
     setIsResponding(isResponding)
     isRespondingRef.current = isResponding
   }, [])
 
-  const getIntroduction = useCallback((str: string) => {
-    return processOpeningStatement(str, formSettings?.inputs || {}, formSettings?.inputsForm || [])
-  }, [formSettings?.inputs, formSettings?.inputsForm])
-  useEffect(() => {
-    if (config?.opening_statement) {
-      handleUpdateChatList(produce(chatListRef.current, (draft) => {
-        const index = draft.findIndex(item => item.isOpeningStatement)
-
-        if (index > -1) {
-          draft[index] = {
-            ...draft[index],
-            content: getIntroduction(config.opening_statement),
-            suggestedQuestions: config.suggested_questions,
-          }
-        }
-        else {
-          draft.unshift({
-            id: `${Date.now()}`,
-            content: getIntroduction(config.opening_statement),
-            isAnswer: true,
-            isOpeningStatement: true,
-            suggestedQuestions: config.suggested_questions,
-          })
-        }
-      }))
-    }
-  }, [config?.opening_statement, getIntroduction, config?.suggested_questions, handleUpdateChatList])
-
   const handleStop = useCallback(() => {
     hasStopResponded.current = true
     handleResponding(false)
@@ -123,50 +167,50 @@ export const useChat = (
     conversationId.current = ''
     taskIdRef.current = ''
     handleStop()
-    const newChatList = config?.opening_statement
-      ? [{
-        id: `${Date.now()}`,
-        content: config.opening_statement,
-        isAnswer: true,
-        isOpeningStatement: true,
-        suggestedQuestions: config.suggested_questions,
-      }]
-      : []
-    handleUpdateChatList(newChatList)
+    setChatTree([])
     setSuggestQuestions([])
-  }, [
-    config,
-    handleStop,
-    handleUpdateChatList,
-  ])
+  }, [handleStop])
 
-  const updateCurrentQA = useCallback(({
+  const updateCurrentQAOnTree = useCallback(({
+    parentId,
     responseItem,
-    questionId,
-    placeholderAnswerId,
+    placeholderQuestionId,
     questionItem,
   }: {
+    parentId?: string
     responseItem: ChatItem
-    questionId: string
-    placeholderAnswerId: string
+    placeholderQuestionId: string
     questionItem: ChatItem
   }) => {
-    const newListWithAnswer = produce(
-      chatListRef.current.filter(item => item.id !== responseItem.id && item.id !== placeholderAnswerId),
-      (draft) => {
-        if (!draft.find(item => item.id === questionId))
-          draft.push({ ...questionItem })
-
-        draft.push({ ...responseItem })
+    let nextState: ChatItemInTree[]
+    const currentQA = { ...questionItem, children: [{ ...responseItem, children: [] }] }
+    if (!parentId && !chatTree.some(item => [placeholderQuestionId, questionItem.id].includes(item.id))) {
+      // QA whose parent is not provided is considered as a first message of the conversation,
+      // and it should be a root node of the chat tree
+      nextState = produce(chatTree, (draft) => {
+        draft.push(currentQA)
       })
-    handleUpdateChatList(newListWithAnswer)
-  }, [handleUpdateChatList])
+    }
+    else {
+      // find the target QA in the tree and update it; if not found, insert it to its parent node
+      nextState = produceChatTreeNode(parentId!, (parentNode) => {
+        const questionNodeIndex = parentNode.children!.findIndex(item => [placeholderQuestionId, questionItem.id].includes(item.id))
+        if (questionNodeIndex === -1)
+          parentNode.children!.push(currentQA)
+        else
+          parentNode.children![questionNodeIndex] = currentQA
+      })
+    }
+    setChatTree(nextState)
+    chatTreeRef.current = nextState
+  }, [chatTree, produceChatTreeNode])
 
   const handleSend = useCallback(async (
     url: string,
     data: {
       query: string
       files?: FileEntity[]
+      parent_message_id?: string
       [key: string]: any
     },
     {
@@ -183,12 +227,15 @@ export const useChat = (
       return false
     }
 
-    const questionId = `question-${Date.now()}`
+    const parentMessage = threadMessages.find(item => item.id === data.parent_message_id)
+
+    const placeholderQuestionId = `question-${Date.now()}`
     const questionItem = {
-      id: questionId,
+      id: placeholderQuestionId,
       content: data.query,
       isAnswer: false,
       message_files: data.files,
+      parentMessageId: data.parent_message_id,
     }
 
     const placeholderAnswerId = `answer-placeholder-${Date.now()}`
@@ -196,18 +243,27 @@ export const useChat = (
       id: placeholderAnswerId,
       content: '',
       isAnswer: true,
+      parentMessageId: questionItem.id,
+      siblingIndex: parentMessage?.children?.length ?? chatTree.length,
     }
 
-    const newList = [...chatListRef.current, questionItem, placeholderAnswerItem]
-    handleUpdateChatList(newList)
+    setTargetMessageId(parentMessage?.id)
+    updateCurrentQAOnTree({
+      parentId: data.parent_message_id,
+      responseItem: placeholderAnswerItem,
+      placeholderQuestionId,
+      questionItem,
+    })
 
     // answer
-    const responseItem: ChatItem = {
+    const responseItem: ChatItemInTree = {
       id: placeholderAnswerId,
       content: '',
       agent_thoughts: [],
       message_files: [],
       isAnswer: true,
+      parentMessageId: questionItem.id,
+      siblingIndex: parentMessage?.children?.length ?? chatTree.length,
     }
 
     handleResponding(true)
@@ -268,7 +324,9 @@ export const useChat = (
           }
 
           if (messageId && !hasSetResponseId) {
+            questionItem.id = `question-${messageId}`
             responseItem.id = messageId
+            responseItem.parentMessageId = questionItem.id
             hasSetResponseId = true
           }
 
@@ -279,11 +337,11 @@ export const useChat = (
           if (messageId)
             responseItem.id = messageId
 
-          updateCurrentQA({
-            responseItem,
-            questionId,
-            placeholderAnswerId,
+          updateCurrentQAOnTree({
+            placeholderQuestionId,
             questionItem,
+            responseItem,
+            parentId: data.parent_message_id,
           })
         },
         async onCompleted(hasError?: boolean) {
@@ -304,43 +362,32 @@ export const useChat = (
             if (!newResponseItem)
               return
 
-            const newChatList = produce(chatListRef.current, (draft) => {
-              const index = draft.findIndex(item => item.id === responseItem.id)
-              if (index !== -1) {
-                const question = draft[index - 1]
-                draft[index - 1] = {
-                  ...question,
-                }
-                draft[index] = {
-                  ...draft[index],
-                  content: newResponseItem.answer,
-                  log: [
-                    ...newResponseItem.message,
-                    ...(newResponseItem.message[newResponseItem.message.length - 1].role !== 'assistant'
-                      ? [
-                        {
-                          role: 'assistant',
-                          text: newResponseItem.answer,
-                          files: newResponseItem.message_files?.filter((file: any) => file.belongs_to === 'assistant') || [],
-                        },
-                      ]
-                      : []),
-                  ],
-                  more: {
-                    time: formatTime(newResponseItem.created_at, 'hh:mm A'),
-                    tokens: newResponseItem.answer_tokens + newResponseItem.message_tokens,
-                    latency: newResponseItem.provider_response_latency.toFixed(2),
-                  },
-                  // for agent log
-                  conversationId: conversationId.current,
-                  input: {
-                    inputs: newResponseItem.inputs,
-                    query: newResponseItem.query,
-                  },
-                }
-              }
+            updateChatTreeNode(responseItem.id, {
+              content: newResponseItem.answer,
+              log: [
+                ...newResponseItem.message,
+                ...(newResponseItem.message[newResponseItem.message.length - 1].role !== 'assistant'
+                  ? [
+                    {
+                      role: 'assistant',
+                      text: newResponseItem.answer,
+                      files: newResponseItem.message_files?.filter((file: any) => file.belongs_to === 'assistant') || [],
+                    },
+                  ]
+                  : []),
+              ],
+              more: {
+                time: formatTime(newResponseItem.created_at, 'hh:mm A'),
+                tokens: newResponseItem.answer_tokens + newResponseItem.message_tokens,
+                latency: newResponseItem.provider_response_latency.toFixed(2),
+              },
+              // for agent log
+              conversationId: conversationId.current,
+              input: {
+                inputs: newResponseItem.inputs,
+                query: newResponseItem.query,
+              },
             })
-            handleUpdateChatList(newChatList)
           }
           if (config?.suggested_questions_after_answer?.enabled && !hasStopResponded.current && onGetSuggestedQuestions) {
             try {
@@ -360,11 +407,11 @@ export const useChat = (
           if (lastThought)
             responseItem.agent_thoughts![responseItem.agent_thoughts!.length - 1].message_files = [...(lastThought as any).message_files, file]
 
-          updateCurrentQA({
-            responseItem,
-            questionId,
-            placeholderAnswerId,
+          updateCurrentQAOnTree({
+            placeholderQuestionId,
             questionItem,
+            responseItem,
+            parentId: data.parent_message_id,
           })
         },
         onThought(thought) {
@@ -372,6 +419,7 @@ export const useChat = (
           const response = responseItem as any
           if (thought.message_id && !hasSetResponseId)
             response.id = thought.message_id
+
           if (response.agent_thoughts.length === 0) {
             response.agent_thoughts.push(thought)
           }
@@ -387,11 +435,11 @@ export const useChat = (
               responseItem.agent_thoughts!.push(thought)
             }
           }
-          updateCurrentQA({
-            responseItem,
-            questionId,
-            placeholderAnswerId,
+          updateCurrentQAOnTree({
+            placeholderQuestionId,
             questionItem,
+            responseItem,
+            parentId: data.parent_message_id,
           })
         },
         onMessageEnd: (messageEnd) => {
@@ -401,43 +449,36 @@ export const useChat = (
               id: messageEnd.metadata.annotation_reply.id,
               authorName: messageEnd.metadata.annotation_reply.account.name,
             })
-            const baseState = chatListRef.current.filter(item => item.id !== responseItem.id && item.id !== placeholderAnswerId)
-            const newListWithAnswer = produce(
-              baseState,
-              (draft) => {
-                if (!draft.find(item => item.id === questionId))
-                  draft.push({ ...questionItem })
-
-                draft.push({
-                  ...responseItem,
-                })
-              })
-            handleUpdateChatList(newListWithAnswer)
+            updateCurrentQAOnTree({
+              placeholderQuestionId,
+              questionItem,
+              responseItem,
+              parentId: data.parent_message_id,
+            })
             return
           }
           responseItem.citation = messageEnd.metadata?.retriever_resources || []
           const processedFilesFromResponse = getProcessedFilesFromResponse(messageEnd.files || [])
           responseItem.allFiles = uniqBy([...(responseItem.allFiles || []), ...(processedFilesFromResponse || [])], 'id')
 
-          const newListWithAnswer = produce(
-            chatListRef.current.filter(item => item.id !== responseItem.id && item.id !== placeholderAnswerId),
-            (draft) => {
-              if (!draft.find(item => item.id === questionId))
-                draft.push({ ...questionItem })
-
-              draft.push({ ...responseItem })
-            })
-          handleUpdateChatList(newListWithAnswer)
+          updateCurrentQAOnTree({
+            placeholderQuestionId,
+            questionItem,
+            responseItem,
+            parentId: data.parent_message_id,
+          })
         },
         onMessageReplace: (messageReplace) => {
           responseItem.content = messageReplace.answer
         },
         onError() {
           handleResponding(false)
-          const newChatList = produce(chatListRef.current, (draft) => {
-            draft.splice(draft.findIndex(item => item.id === placeholderAnswerId), 1)
+          updateCurrentQAOnTree({
+            placeholderQuestionId,
+            questionItem,
+            responseItem,
+            parentId: data.parent_message_id,
           })
-          handleUpdateChatList(newChatList)
         },
         onWorkflowStarted: ({ workflow_run_id, task_id }) => {
           taskIdRef.current = task_id
@@ -446,89 +487,84 @@ export const useChat = (
             status: WorkflowRunningStatus.Running,
             tracing: [],
           }
-          handleUpdateChatList(produce(chatListRef.current, (draft) => {
-            const currentIndex = draft.findIndex(item => item.id === responseItem.id)
-            draft[currentIndex] = {
-              ...draft[currentIndex],
-              ...responseItem,
-            }
-          }))
+          updateCurrentQAOnTree({
+            placeholderQuestionId,
+            questionItem,
+            responseItem,
+            parentId: data.parent_message_id,
+          })
         },
-        onWorkflowFinished: ({ data }) => {
-          responseItem.workflowProcess!.status = data.status as WorkflowRunningStatus
-          handleUpdateChatList(produce(chatListRef.current, (draft) => {
-            const currentIndex = draft.findIndex(item => item.id === responseItem.id)
-            draft[currentIndex] = {
-              ...draft[currentIndex],
-              ...responseItem,
-            }
-          }))
+        onWorkflowFinished: ({ data: workflowFinishedData }) => {
+          responseItem.workflowProcess!.status = workflowFinishedData.status as WorkflowRunningStatus
+          updateCurrentQAOnTree({
+            placeholderQuestionId,
+            questionItem,
+            responseItem,
+            parentId: data.parent_message_id,
+          })
         },
-        onIterationStart: ({ data }) => {
+        onIterationStart: ({ data: iterationStartedData }) => {
           responseItem.workflowProcess!.tracing!.push({
-            ...data,
+            ...iterationStartedData,
             status: WorkflowRunningStatus.Running,
           } as any)
-          handleUpdateChatList(produce(chatListRef.current, (draft) => {
-            const currentIndex = draft.findIndex(item => item.id === responseItem.id)
-            draft[currentIndex] = {
-              ...draft[currentIndex],
-              ...responseItem,
-            }
-          }))
+          updateCurrentQAOnTree({
+            placeholderQuestionId,
+            questionItem,
+            responseItem,
+            parentId: data.parent_message_id,
+          })
         },
-        onIterationFinish: ({ data }) => {
+        onIterationFinish: ({ data: iterationFinishedData }) => {
           const tracing = responseItem.workflowProcess!.tracing!
-          const iterationIndex = tracing.findIndex(item => item.node_id === data.node_id
-            && (item.execution_metadata?.parallel_id === data.execution_metadata?.parallel_id || item.parallel_id === data.execution_metadata?.parallel_id))!
+          const iterationIndex = tracing.findIndex(item => item.node_id === iterationFinishedData.node_id
+            && (item.execution_metadata?.parallel_id === iterationFinishedData.execution_metadata?.parallel_id || item.parallel_id === iterationFinishedData.execution_metadata?.parallel_id))!
           tracing[iterationIndex] = {
             ...tracing[iterationIndex],
-            ...data,
+            ...iterationFinishedData,
             status: WorkflowRunningStatus.Succeeded,
           } as any
 
-          handleUpdateChatList(produce(chatListRef.current, (draft) => {
-            const currentIndex = draft.findIndex(item => item.id === responseItem.id)
-            draft[currentIndex] = {
-              ...draft[currentIndex],
-              ...responseItem,
-            }
-          }))
+          updateCurrentQAOnTree({
+            placeholderQuestionId,
+            questionItem,
+            responseItem,
+            parentId: data.parent_message_id,
+          })
         },
-        onNodeStarted: ({ data }) => {
-          if (data.iteration_id)
+        onNodeStarted: ({ data: nodeStartedData }) => {
+          if (nodeStartedData.iteration_id)
             return
 
           responseItem.workflowProcess!.tracing!.push({
-            ...data,
+            ...nodeStartedData,
             status: WorkflowRunningStatus.Running,
           } as any)
-          handleUpdateChatList(produce(chatListRef.current, (draft) => {
-            const currentIndex = draft.findIndex(item => item.id === responseItem.id)
-            draft[currentIndex] = {
-              ...draft[currentIndex],
-              ...responseItem,
-            }
-          }))
+          updateCurrentQAOnTree({
+            placeholderQuestionId,
+            questionItem,
+            responseItem,
+            parentId: data.parent_message_id,
+          })
         },
-        onNodeFinished: ({ data }) => {
-          if (data.iteration_id)
+        onNodeFinished: ({ data: nodeFinishedData }) => {
+          if (nodeFinishedData.iteration_id)
             return
 
           const currentIndex = responseItem.workflowProcess!.tracing!.findIndex((item) => {
             if (!item.execution_metadata?.parallel_id)
-              return item.node_id === data.node_id
+              return item.node_id === nodeFinishedData.node_id
 
-            return item.node_id === data.node_id && (item.execution_metadata?.parallel_id === data.execution_metadata.parallel_id)
+            return item.node_id === nodeFinishedData.node_id && (item.execution_metadata?.parallel_id === nodeFinishedData.execution_metadata.parallel_id)
+          })
+          responseItem.workflowProcess!.tracing[currentIndex] = nodeFinishedData as any
+
+          updateCurrentQAOnTree({
+            placeholderQuestionId,
+            questionItem,
+            responseItem,
+            parentId: data.parent_message_id,
           })
-          responseItem.workflowProcess!.tracing[currentIndex] = data as any
-          handleUpdateChatList(produce(chatListRef.current, (draft) => {
-            const currentIndex = draft.findIndex(item => item.id === responseItem.id)
-            draft[currentIndex] = {
-              ...draft[currentIndex],
-              ...responseItem,
-            }
-          }))
         },
         onTTSChunk: (messageId: string, audio: string) => {
           if (!audio || audio === '')
@@ -542,11 +578,13 @@ export const useChat = (
       })
     return true
   }, [
-    config?.suggested_questions_after_answer,
-    updateCurrentQA,
     t,
+    chatTree.length,
+    threadMessages,
+    config?.suggested_questions_after_answer,
+    updateCurrentQAOnTree,
+    updateChatTreeNode,
     notify,
-    handleUpdateChatList,
     handleResponding,
     formatTime,
     params.token,
@@ -556,76 +594,61 @@ export const useChat = (
   ])
 
   const handleAnnotationEdited = useCallback((query: string, answer: string, index: number) => {
-    handleUpdateChatList(chatListRef.current.map((item, i) => {
-      if (i === index - 1) {
-        return {
-          ...item,
-          content: query,
-        }
-      }
-      if (i === index) {
-        return {
-          ...item,
-          content: answer,
-          annotation: {
-            ...item.annotation,
-            logAnnotation: undefined,
-          } as any,
-        }
-      }
-      return item
-    }))
-  }, [handleUpdateChatList])
+    const targetQuestionId = chatList[index - 1].id
+    const targetAnswerId = chatList[index].id
+
+    updateChatTreeNode(targetQuestionId, {
+      content: query,
+    })
+    updateChatTreeNode(targetAnswerId, {
+      content: answer,
+      annotation: {
+        ...chatList[index].annotation,
+        logAnnotation: undefined,
+      } as any,
+    })
+  }, [chatList, updateChatTreeNode])
+
   const handleAnnotationAdded = useCallback((annotationId: string, authorName: string, query: string, answer: string, index: number) => {
-    handleUpdateChatList(chatListRef.current.map((item, i) => {
-      if (i === index - 1) {
-        return {
-          ...item,
-          content: query,
-        }
-      }
-      if (i === index) {
-        const answerItem = {
-          ...item,
-          content: item.content,
-          annotation: {
-            id: annotationId,
-            authorName,
-            logAnnotation: {
-              content: answer,
-              account: {
-                id: '',
-                name: authorName,
-                email: '',
-              },
-            },
-          } as Annotation,
-        }
-        return answerItem
-      }
-      return item
-    }))
-  }, [handleUpdateChatList])
-  const handleAnnotationRemoved = useCallback((index: number) => {
-    handleUpdateChatList(chatListRef.current.map((item, i) => {
-      if (i === index) {
-        return {
-          ...item,
-          content: item.content,
-          annotation: {
-            ...(item.annotation || {}),
+    const targetQuestionId = chatList[index - 1].id
+    const targetAnswerId = chatList[index].id
+
+    updateChatTreeNode(targetQuestionId, {
+      content: query,
+    })
+
+    updateChatTreeNode(targetAnswerId, {
+      content: chatList[index].content,
+      annotation: {
+        id: annotationId,
+        authorName,
+        logAnnotation: {
+          content: answer,
+          account: {
             id: '',
-          } as Annotation,
-        }
-      }
-      return item
-    }))
-  }, [handleUpdateChatList])
+            name: authorName,
+            email: '',
+          },
+        },
+      } as Annotation,
+    })
+  }, [chatList, updateChatTreeNode])
+
+  const handleAnnotationRemoved = useCallback((index: number) => {
+    const targetAnswerId = chatList[index].id
+
+    updateChatTreeNode(targetAnswerId, {
+      content: chatList[index].content,
+      annotation: {
+        ...(chatList[index].annotation || {}),
+        id: '',
+      } as Annotation,
+    })
+  }, [chatList, updateChatTreeNode])
 
   return {
     chatList,
-    chatListRef,
-    handleUpdateChatList,
+    setTargetMessageId,
     conversationId: conversationId.current,
     isResponding,
     setIsResponding,

web/app/components/base/chat/embedded-chatbot/chat-wrapper.tsx

@@ -3,10 +3,11 @@ import Chat from '../chat'
 import type {
   ChatConfig,
   ChatItem,
+  ChatItemInTree,
   OnSend,
 } from '../types'
 import { useChat } from '../chat/hooks'
-import { getLastAnswer } from '../utils'
+import { getLastAnswer, isValidGeneratedAnswer } from '../utils'
 import { useEmbeddedChatbotContext } from './context'
 import ConfigPanel from './config-panel'
 import { isDify } from './utils'
@@ -51,13 +52,12 @@ const ChatWrapper = () => {
     } as ChatConfig
   }, [appParams, currentConversationItem?.introduction, currentConversationId])
   const {
-    chatListRef,
     chatList,
+    setTargetMessageId,
     handleSend,
     handleStop,
     isResponding,
     suggestedQuestions,
-    handleUpdateChatList,
   } = useChat(
     appConfig,
     {
@@ -71,15 +71,15 @@ const ChatWrapper = () => {
   useEffect(() => {
     if (currentChatInstanceRef.current)
       currentChatInstanceRef.current.handleStop = handleStop
-  }, [])
+  }, [currentChatInstanceRef, handleStop])
 
-  const doSend: OnSend = useCallback((message, files, last_answer) => {
+  const doSend: OnSend = useCallback((message, files, isRegenerate = false, parentAnswer: ChatItem | null = null) => {
     const data: any = {
       query: message,
       files,
       inputs: currentConversationId ? currentConversationItem?.inputs : newConversationInputs,
       conversation_id: currentConversationId,
-      parent_message_id: last_answer?.id || getLastAnswer(chatListRef.current)?.id || null,
+      parent_message_id: (isRegenerate ? parentAnswer?.id : getLastAnswer(chatList)?.id) || null,
     }
 
     handleSend(
@@ -92,32 +92,21 @@ const ChatWrapper = () => {
       },
     )
   }, [
-    chatListRef,
-    appConfig,
+    chatList,
+    handleNewConversationCompleted,
+    handleSend,
     currentConversationId,
     currentConversationItem,
-    handleSend,
     newConversationInputs,
-    handleNewConversationCompleted,
     isInstalledApp,
     appId,
   ])
 
-  const doRegenerate = useCallback((chatItem: ChatItem) => {
-    const index = chatList.findIndex(item => item.id === chatItem.id)
-    if (index === -1)
-      return
-
-    const prevMessages = chatList.slice(0, index)
-    const question = prevMessages.pop()
-    const lastAnswer = getLastAnswer(prevMessages)
-
-    if (!question)
-      return
-
-    handleUpdateChatList(prevMessages)
-    doSend(question.content, question.message_files, lastAnswer)
-  }, [chatList, handleUpdateChatList, doSend])
+  const doRegenerate = useCallback((chatItem: ChatItemInTree) => {
+    const question = chatList.find(item => item.id === chatItem.parentMessageId)!
+    const parentAnswer = chatList.find(item => item.id === question.parentMessageId)
+    doSend(question.content, question.message_files, true, isValidGeneratedAnswer(parentAnswer) ? parentAnswer : null)
+  }, [chatList, doSend])
 
   const chatNode = useMemo(() => {
     if (inputsForms.length) {
@@ -172,6 +161,7 @@ const ChatWrapper = () => {
       answerIcon={answerIcon}
       hideProcessDetail
       themeBuilder={themeBuilder}
+      switchSibling={siblingMessageId => setTargetMessageId(siblingMessageId)}
     />
   )
 }

web/app/components/base/chat/types.ts

@@ -67,9 +67,12 @@ export type ChatItem = IChatItem & {
 
 export type ChatItemInTree = {
   children?: ChatItemInTree[]
-} & IChatItem
+} & ChatItem
 
-export type OnSend = (message: string, files?: FileEntity[], last_answer?: ChatItem | null) => void
+export type OnSend = {
+  (message: string, files?: FileEntity[]): void
+  (message: string, files: FileEntity[] | undefined, isRegenerate: boolean, lastAnswer?: ChatItem | null): void
+}
 
 export type OnRegenerate = (chatItem: ChatItem) => void
 

web/app/components/base/chat/utils.ts

@@ -1,8 +1,6 @@
-import { addFileInfos, sortAgentSorts } from '../../tools/utils'
 import { UUID_NIL } from './constants'
 import type { IChatItem } from './chat/type'
 import type { ChatItem, ChatItemInTree } from './types'
-import { getProcessedFilesFromResponse } from '@/app/components/base/file-uploader/utils'
 
 async function decodeBase64AndDecompress(base64String: string) {
   const binaryString = atob(base64String)
@@ -21,67 +19,24 @@ function getProcessedInputsFromUrlParams(): Record<string, any> {
   return inputs
 }
 
-function getLastAnswer(chatList: ChatItem[]) {
+function isValidGeneratedAnswer(item?: ChatItem | ChatItemInTree): boolean {
+  return !!item && item.isAnswer && !item.id.startsWith('answer-placeholder-') && !item.isOpeningStatement
+}
+
+function getLastAnswer<T extends ChatItem | ChatItemInTree>(chatList: T[]): T | null {
   for (let i = chatList.length - 1; i >= 0; i--) {
     const item = chatList[i]
-    if (item.isAnswer && !item.id.startsWith('answer-placeholder-') && !item.isOpeningStatement)
+    if (isValidGeneratedAnswer(item))
       return item
   }
   return null
 }
 
-function appendQAToChatList(chatList: ChatItem[], item: any) {
-  // we append answer first and then question since will reverse the whole chatList later
-  const answerFiles = item.message_files?.filter((file: any) => file.belongs_to === 'assistant') || []
-  chatList.push({
-    id: item.id,
-    content: item.answer,
-    agent_thoughts: addFileInfos(item.agent_thoughts ? sortAgentSorts(item.agent_thoughts) : item.agent_thoughts, item.message_files),
-    feedback: item.feedback,
-    isAnswer: true,
-    citation: item.retriever_resources,
-    message_files: getProcessedFilesFromResponse(answerFiles.map((item: any) => ({ ...item, related_id: item.id }))),
-  })
-  const questionFiles = item.message_files?.filter((file: any) => file.belongs_to === 'user') || []
-  chatList.push({
-    id: `question-${item.id}`,
-    content: item.query,
-    isAnswer: false,
-    message_files: getProcessedFilesFromResponse(questionFiles.map((item: any) => ({ ...item, related_id: item.id }))),
-  })
-}
-
 /**
- * Computes the latest thread messages from all messages of the conversation.
- * Same logic as backend codebase `api/core/prompt/utils/extract_thread_messages.py`
- *
- * @param fetchedMessages - The history chat list data from the backend, sorted by created_at in descending order. This includes all flattened history messages of the conversation.
- * @returns An array of ChatItems representing the latest thread.
+ * Build a chat item tree from a chat list
+ * @param allMessages - The chat list, sorted from oldest to newest
+ * @returns The chat item tree
  */
-function getPrevChatList(fetchedMessages: any[]) {
-  const ret: ChatItem[] = []
-  let nextMessageId = null
-
-  for (const item of fetchedMessages) {
-    if (!item.parent_message_id) {
-      appendQAToChatList(ret, item)
-      break
-    }
-
-    if (!nextMessageId) {
-      appendQAToChatList(ret, item)
-      nextMessageId = item.parent_message_id
-    }
-    else {
-      if (item.id === nextMessageId || nextMessageId === UUID_NIL) {
-        appendQAToChatList(ret, item)
-        nextMessageId = item.parent_message_id
-      }
-    }
-  }
-  return ret.reverse()
-}
-
 function buildChatItemTree(allMessages: IChatItem[]): ChatItemInTree[] {
   const map: Record<string, ChatItemInTree> = {}
   const rootNodes: ChatItemInTree[] = []
@@ -208,7 +163,7 @@ function getThreadMessages(tree: ChatItemInTree[], targetMessageId?: string): Ch
 
 export {
   getProcessedInputsFromUrlParams,
-  getPrevChatList,
+  isValidGeneratedAnswer,
   getLastAnswer,
   buildChatItemTree,
   getThreadMessages,

web/app/components/base/markdown.tsx

@@ -229,7 +229,11 @@ export function Markdown(props: { content: string; className?: string }) {
   return (
     <div className={cn(props.className, 'markdown-body')}>
       <ReactMarkdown
-        remarkPlugins={[RemarkGfm, RemarkMath, RemarkBreaks]}
+        remarkPlugins={[
+          RemarkGfm,
+          [RemarkMath, { singleDollarTextMath: false }],
+          RemarkBreaks,
+        ]}
         rehypePlugins={[
           RehypeKatex,
           RehypeRaw as any,

web/app/components/base/prompt-editor/index.tsx

@@ -149,7 +149,7 @@ const PromptEditor: FC<PromptEditorProps> = ({
     <LexicalComposer initialConfig={{ ...initialConfig, editable }}>
       <div className='relative min-h-5'>
         <RichTextPlugin
-          contentEditable={<ContentEditable className={`${className} outline-none ${compact ? 'leading-5 text-[13px]' : 'leading-6 text-sm'} text-gray-700`} style={style || {}} />}
+          contentEditable={<ContentEditable className={`${className} outline-none ${compact ? 'leading-5 text-[13px]' : 'leading-6 text-sm'} text-text-secondary`} style={style || {}} />}
           placeholder={<Placeholder value={placeholder} className={cn('truncate', placeholderClassName)} compact={compact} />}
           ErrorBoundary={LexicalErrorBoundary}
         />

web/app/components/base/prompt-editor/plugins/component-picker-block/hooks.tsx

@@ -133,7 +133,7 @@ export const useVariableOptions = (
           return (
             <VariableMenuItem
               title={item.value}
-              icon={<BracketsX className='w-[14px] h-[14px] text-[#2970FF]' />}
+              icon={<BracketsX className='w-[14px] h-[14px] text-text-accent' />}
               queryString={queryString}
               isSelected={isSelected}
               onClick={onSelect}
@@ -162,7 +162,7 @@ export const useVariableOptions = (
         return (
           <VariableMenuItem
             title={t('common.promptEditor.variable.modal.add')}
-            icon={<BracketsX className='mr-2 w-[14px] h-[14px] text-[#2970FF]' />}
+            icon={<BracketsX className='w-[14px] h-[14px] text-text-accent' />}
             queryString={queryString}
             isSelected={isSelected}
             onClick={onSelect}
@@ -211,7 +211,7 @@ export const useExternalToolOptions = (
                   background={item.icon_background}
                 />
               }
-              extraElement={<div className='text-xs text-gray-400'>{item.variableName}</div>}
+              extraElement={<div className='text-xs text-text-tertiary'>{item.variableName}</div>}
               queryString={queryString}
               isSelected={isSelected}
               onClick={onSelect}
@@ -240,8 +240,8 @@ export const useExternalToolOptions = (
         return (
           <VariableMenuItem
             title={t('common.promptEditor.variable.modal.addTool')}
-            icon={<Tool03 className='mr-2 w-[14px] h-[14px] text-[#444CE7]' />}
-            extraElement={< ArrowUpRight className='w-3 h-3 text-gray-400' />}
+            icon={<Tool03 className='w-[14px] h-[14px] text-text-accent' />}
+            extraElement={< ArrowUpRight className='w-3 h-3 text-text-tertiary' />}
             queryString={queryString}
             isSelected={isSelected}
             onClick={onSelect}

web/app/components/base/prompt-editor/plugins/component-picker-block/index.tsx

@@ -135,7 +135,7 @@ const ComponentPicker = ({
             // See https://github.com/facebook/lexical/blob/ac97dfa9e14a73ea2d6934ff566282d7f758e8bb/packages/lexical-react/src/shared/LexicalMenu.ts#L493
             <div className='w-0 h-0'>
               <div
-                className='p-1 w-[260px] bg-white rounded-lg border-[0.5px] border-gray-200 shadow-lg'
+                className='p-1 w-[260px] bg-components-panel-bg-blur rounded-lg border-[0.5px] border-components-panel-border shadow-lg'
                 style={{
                   ...floatingStyles,
                   visibility: isPositioned ? 'visible' : 'hidden',
@@ -148,7 +148,7 @@ const ComponentPicker = ({
                       {
                         // Divider
                         index !== 0 && options.at(index - 1)?.group !== option.group && (
-                          <div className='h-px bg-gray-100 my-1 w-full -translate-x-1'></div>
+                          <div className='h-px bg-divider-subtle my-1 w-full -translate-x-1'></div>
                         )
                       }
                       {option.renderMenuOption({
@@ -169,7 +169,7 @@ const ComponentPicker = ({
                     <>
                       {
                         (!!options.length) && (
-                          <div className='h-px bg-gray-100 my-1 w-full -translate-x-1'></div>
+                          <div className='h-px bg-divider-subtle my-1 w-full -translate-x-1'></div>
                         )
                       }
                       <div className='p-1'>

web/app/components/base/prompt-editor/plugins/component-picker-block/prompt-option.tsx

@@ -21,9 +21,9 @@ export const PromptMenuItem = memo(({
   return (
     <div
       className={`
-        flex items-center px-3 h-6 cursor-pointer hover:bg-gray-50 rounded-md
-        ${isSelected && !disabled && '!bg-gray-50'}
-        ${disabled ? 'cursor-not-allowed opacity-30' : 'hover:bg-gray-50 cursor-pointer'}
+        flex items-center px-3 h-6 cursor-pointer hover:bg-state-base-hover rounded-md
+        ${isSelected && !disabled && '!bg-state-base-hover'}
+        ${disabled ? 'cursor-not-allowed opacity-30' : 'hover:bg-state-base-hover cursor-pointer'}
       `}
       tabIndex={-1}
       ref={setRefElement}
@@ -38,7 +38,7 @@ export const PromptMenuItem = memo(({
         onClick()
       }}>
       {icon}
-      <div className='ml-1 text-[13px] text-gray-900'>{title}</div>
+      <div className='ml-1 text-[13px] text-text-secondary'>{title}</div>
     </div>
   )
 })

web/app/components/base/prompt-editor/plugins/component-picker-block/variable-option.tsx

@@ -38,8 +38,8 @@ export const VariableMenuItem = memo(({
   return (
     <div
       className={`
-        flex items-center px-3 h-6 rounded-md hover:bg-primary-50 cursor-pointer
-        ${isSelected && 'bg-primary-50'}
+        flex items-center px-3 h-6 rounded-md hover:bg-state-base-hover cursor-pointer
+        ${isSelected && 'bg-state-base-hover'}
       `}
       tabIndex={-1}
       ref={setRefElement}
@@ -48,9 +48,9 @@ export const VariableMenuItem = memo(({
       <div className='mr-2'>
         {icon}
       </div>
-      <div className='grow text-[13px] text-gray-900 truncate' title={title}>
+      <div className='grow text-[13px] text-text-secondary truncate' title={title}>
         {before}
-        <span className='text-[#2970FF]'>{middle}</span>
+        <span className='text-text-accent'>{middle}</span>
         {after}
       </div>
       {extraElement}

web/app/components/base/prompt-editor/plugins/placeholder.tsx

@@ -16,7 +16,7 @@ const Placeholder = ({
   return (
     <div className={cn(
       className,
-      'absolute top-0 left-0 h-full w-full text-sm text-gray-300 select-none pointer-events-none',
+      'absolute top-0 left-0 h-full w-full text-sm text-components-input-text-placeholder select-none pointer-events-none',
       compact ? 'leading-5 text-[13px]' : 'leading-6 text-sm',
     )}>
       {value || t('common.promptEditor.placeholder')}

web/app/components/base/prompt-editor/plugins/variable-value-block/node.tsx

@@ -24,7 +24,7 @@ export class VariableValueBlockNode extends TextNode {
 
   createDOM(config: EditorConfig): HTMLElement {
     const element = super.createDOM(config)
-    element.classList.add('inline-flex', 'items-center', 'px-0.5', 'h-[22px]', 'text-[#155EEF]', 'rounded-[5px]', 'align-middle')
+    element.classList.add('inline-flex', 'items-center', 'px-0.5', 'h-[22px]', 'text-text-accent', 'rounded-[5px]', 'align-middle')
     return element
   }