chore: Add PyPI package support for uvx/pip installation (#1764)

* Initial plan * Add package structure and resource path utilities - Created langbot/ package with __init__.py and __main__.py entry point - Added paths utility to find frontend and resource files from package installation - Updated config loading to use resource paths - Updated frontend serving to use resource paths - Added MANIFEST.in for package data inclusion - Updated pyproject.toml with build system and entry points Co-authored-by: RockChinQ <45992437+RockChinQ@users.noreply.github.com> * Add PyPI publishing workflow and update license - Created GitHub Actions workflow to build frontend and publish to PyPI - Added license field to pyproject.toml to fix deprecation warning - Updated .gitignore to exclude build artifacts - Tested package building successfully Co-authored-by: RockChinQ <45992437+RockChinQ@users.noreply.github.com> * Add PyPI installation documentation - Created PYPI_INSTALLATION.md with detailed installation and usage instructions - Updated README.md to feature uvx/pip installation as recommended method - Updated README_EN.md with same changes for English documentation Co-authored-by: RockChinQ <45992437+RockChinQ@users.noreply.github.com> * Address code review feedback - Made package-data configuration more specific to langbot package only - Improved path detection with caching to avoid repeated file I/O - Removed sys.path searching which was incorrect for package data - Removed interactive input() call for non-interactive environment compatibility - Simplified error messages for version check Co-authored-by: RockChinQ <45992437+RockChinQ@users.noreply.github.com> * Fix code review issues - Use specific exception types instead of bare except - Fix misleading comments about directory levels - Remove redundant existence check before makedirs with exist_ok=True - Use context manager for file opening to ensure proper cleanup Co-authored-by: RockChinQ <45992437+RockChinQ@users.noreply.github.com> * Simplify package configuration and document behavioral differences - Removed redundant package-data configuration, relying on MANIFEST.in - Added documentation about behavioral differences between package and source installation - Clarified that include-package-data=true uses MANIFEST.in for data files Co-authored-by: RockChinQ <45992437+RockChinQ@users.noreply.github.com> * chore: update pyproject.toml * chore: try pack templates in langbot/ * chore: update * chore: update * chore: update * chore: update * chore: update * chore: adjust dir structure * chore: fix imports * fix: read default-pipeline-config.json * fix: read default-pipeline-config.json * fix: tests * ci: publish pypi * chore: bump version 4.6.0-beta.1 for testing * chore: add templates/** * fix: send adapters and requesters icons * chore: bump version 4.6.0b2 for testing * chore: add platform field for docker-compose.yaml --------- Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com> Co-authored-by: RockChinQ <45992437+RockChinQ@users.noreply.github.com> Co-authored-by: Junyan Qin <rockchinq@gmail.com>
2026-06-12 00:36:03 +00:00 · 2025-11-16 19:53:01 +08:00
parent 6a24c951e0
commit e642ffa5b3
477 changed files with 1001 additions and 1002 deletions
--- a/src/langbot/pkg/provider/modelmgr/init.py
+++ b/src/langbot/pkg/provider/modelmgr/init.py
--- a/src/langbot/pkg/provider/modelmgr/entities.py
+++ b/src/langbot/pkg/provider/modelmgr/entities.py
@@ -0,0 +1,27 @@
+from __future__ import annotations
+
+import typing
+
+import pydantic
+
+from . import requester
+from . import token
+
+
+class LLMModelInfo(pydantic.BaseModel):
+    """模型"""
+
+    name: str
+
+    model_name: typing.Optional[str] = None
+
+    token_mgr: token.TokenManager
+
+    requester: requester.ProviderAPIRequester
+
+    tool_call_supported: typing.Optional[bool] = False
+
+    vision_supported: typing.Optional[bool] = False
+
+    class Config:
+        arbitrary_types_allowed = True
--- a/src/langbot/pkg/provider/modelmgr/errors.py
+++ b/src/langbot/pkg/provider/modelmgr/errors.py
@@ -0,0 +1,5 @@
+class RequesterError(Exception):
+    """Base class for all Requester errors."""
+
+    def __init__(self, message: str):
+        super().__init__('模型请求失败: ' + message)
--- a/src/langbot/pkg/provider/modelmgr/modelmgr.py
+++ b/src/langbot/pkg/provider/modelmgr/modelmgr.py
@@ -0,0 +1,202 @@
+from __future__ import annotations
+
+import sqlalchemy
+import traceback
+
+from . import requester
+from ...core import app
+from ...discover import engine
+from . import token
+from ...entity.persistence import model as persistence_model
+from ...entity.errors import provider as provider_errors
+
+FETCH_MODEL_LIST_URL = 'https://api.qchatgpt.rockchin.top/api/v2/fetch/model_list'
+
+
+class ModelManager:
+    """模型管理器"""
+
+    ap: app.Application
+
+    llm_models: list[requester.RuntimeLLMModel]
+
+    embedding_models: list[requester.RuntimeEmbeddingModel]
+
+    requester_components: list[engine.Component]
+
+    requester_dict: dict[str, type[requester.ProviderAPIRequester]]  # cache
+
+    def __init__(self, ap: app.Application):
+        self.ap = ap
+        self.llm_models = []
+        self.embedding_models = []
+        self.requester_components = []
+        self.requester_dict = {}
+
+    async def initialize(self):
+        self.requester_components = self.ap.discover.get_components_by_kind('LLMAPIRequester')
+
+        # forge requester class dict
+        requester_dict: dict[str, type[requester.ProviderAPIRequester]] = {}
+        for component in self.requester_components:
+            requester_dict[component.metadata.name] = component.get_python_component_class()
+
+        self.requester_dict = requester_dict
+
+        await self.load_models_from_db()
+
+    async def load_models_from_db(self):
+        """从数据库加载模型"""
+        self.ap.logger.info('Loading models from db...')
+
+        self.llm_models = []
+        self.embedding_models = []
+
+        # llm models
+        result = await self.ap.persistence_mgr.execute_async(sqlalchemy.select(persistence_model.LLMModel))
+        llm_models = result.all()
+        for llm_model in llm_models:
+            try:
+                await self.load_llm_model(llm_model)
+            except provider_errors.RequesterNotFoundError as e:
+                self.ap.logger.warning(f'Requester {e.requester_name} not found, skipping llm model {llm_model.uuid}')
+            except Exception as e:
+                self.ap.logger.error(f'Failed to load model {llm_model.uuid}: {e}\n{traceback.format_exc()}')
+
+        # embedding models
+        result = await self.ap.persistence_mgr.execute_async(sqlalchemy.select(persistence_model.EmbeddingModel))
+        embedding_models = result.all()
+        for embedding_model in embedding_models:
+            try:
+                await self.load_embedding_model(embedding_model)
+            except provider_errors.RequesterNotFoundError as e:
+                self.ap.logger.warning(
+                    f'Requester {e.requester_name} not found, skipping embedding model {embedding_model.uuid}'
+                )
+            except Exception as e:
+                self.ap.logger.error(f'Failed to load model {embedding_model.uuid}: {e}\n{traceback.format_exc()}')
+
+    async def init_runtime_llm_model(
+        self,
+        model_info: persistence_model.LLMModel | sqlalchemy.Row[persistence_model.LLMModel] | dict,
+    ):
+        """初始化运行时 LLM 模型"""
+        if isinstance(model_info, sqlalchemy.Row):
+            model_info = persistence_model.LLMModel(**model_info._mapping)
+        elif isinstance(model_info, dict):
+            model_info = persistence_model.LLMModel(**model_info)
+
+        if model_info.requester not in self.requester_dict:
+            raise provider_errors.RequesterNotFoundError(model_info.requester)
+
+        requester_inst = self.requester_dict[model_info.requester](ap=self.ap, config=model_info.requester_config)
+
+        await requester_inst.initialize()
+
+        runtime_llm_model = requester.RuntimeLLMModel(
+            model_entity=model_info,
+            token_mgr=token.TokenManager(
+                name=model_info.uuid,
+                tokens=model_info.api_keys,
+            ),
+            requester=requester_inst,
+        )
+
+        return runtime_llm_model
+
+    async def init_runtime_embedding_model(
+        self,
+        model_info: persistence_model.EmbeddingModel | sqlalchemy.Row[persistence_model.EmbeddingModel] | dict,
+    ):
+        """初始化运行时 Embedding 模型"""
+        if isinstance(model_info, sqlalchemy.Row):
+            model_info = persistence_model.EmbeddingModel(**model_info._mapping)
+        elif isinstance(model_info, dict):
+            model_info = persistence_model.EmbeddingModel(**model_info)
+
+        if model_info.requester not in self.requester_dict:
+            raise provider_errors.RequesterNotFoundError(model_info.requester)
+
+        requester_inst = self.requester_dict[model_info.requester](ap=self.ap, config=model_info.requester_config)
+
+        await requester_inst.initialize()
+
+        runtime_embedding_model = requester.RuntimeEmbeddingModel(
+            model_entity=model_info,
+            token_mgr=token.TokenManager(
+                name=model_info.uuid,
+                tokens=model_info.api_keys,
+            ),
+            requester=requester_inst,
+        )
+
+        return runtime_embedding_model
+
+    async def load_llm_model(
+        self,
+        model_info: persistence_model.LLMModel | sqlalchemy.Row[persistence_model.LLMModel] | dict,
+    ):
+        """加载 LLM 模型"""
+        runtime_llm_model = await self.init_runtime_llm_model(model_info)
+        self.llm_models.append(runtime_llm_model)
+
+    async def load_embedding_model(
+        self,
+        model_info: persistence_model.EmbeddingModel | sqlalchemy.Row[persistence_model.EmbeddingModel] | dict,
+    ):
+        """加载 Embedding 模型"""
+        runtime_embedding_model = await self.init_runtime_embedding_model(model_info)
+        self.embedding_models.append(runtime_embedding_model)
+
+    async def get_model_by_uuid(self, uuid: str) -> requester.RuntimeLLMModel:
+        """通过uuid获取 LLM 模型"""
+        for model in self.llm_models:
+            if model.model_entity.uuid == uuid:
+                return model
+        raise ValueError(f'LLM model {uuid} not found')
+
+    async def get_embedding_model_by_uuid(self, uuid: str) -> requester.RuntimeEmbeddingModel:
+        """通过uuid获取 Embedding 模型"""
+        for model in self.embedding_models:
+            if model.model_entity.uuid == uuid:
+                return model
+        raise ValueError(f'Embedding model {uuid} not found')
+
+    async def remove_llm_model(self, model_uuid: str):
+        """移除 LLM 模型"""
+        for model in self.llm_models:
+            if model.model_entity.uuid == model_uuid:
+                self.llm_models.remove(model)
+                return
+
+    async def remove_embedding_model(self, model_uuid: str):
+        """移除 Embedding 模型"""
+        for model in self.embedding_models:
+            if model.model_entity.uuid == model_uuid:
+                self.embedding_models.remove(model)
+                return
+
+    def get_available_requesters_info(self, model_type: str) -> list[dict]:
+        """获取所有可用的请求器"""
+        if model_type != '':
+            return [
+                component.to_plain_dict()
+                for component in self.requester_components
+                if model_type in component.spec['support_type']
+            ]
+        else:
+            return [component.to_plain_dict() for component in self.requester_components]
+
+    def get_available_requester_info_by_name(self, name: str) -> dict | None:
+        """通过名称获取请求器信息"""
+        for component in self.requester_components:
+            if component.metadata.name == name:
+                return component.to_plain_dict()
+        return None
+
+    def get_available_requester_manifest_by_name(self, name: str) -> engine.Component | None:
+        """通过名称获取请求器清单"""
+        for component in self.requester_components:
+            if component.metadata.name == name:
+                return component
+        return None
--- a/src/langbot/pkg/provider/modelmgr/requester.py
+++ b/src/langbot/pkg/provider/modelmgr/requester.py
@@ -0,0 +1,142 @@
+from __future__ import annotations
+
+import abc
+import typing
+
+from ...core import app
+from ...entity.persistence import model as persistence_model
+import langbot_plugin.api.entities.builtin.resource.tool as resource_tool
+from . import token
+import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
+import langbot_plugin.api.entities.builtin.provider.message as provider_message
+
+
+class RuntimeLLMModel:
+    """运行时模型"""
+
+    model_entity: persistence_model.LLMModel
+    """模型数据"""
+
+    token_mgr: token.TokenManager
+    """api key管理器"""
+
+    requester: ProviderAPIRequester
+    """请求器实例"""
+
+    def __init__(
+        self,
+        model_entity: persistence_model.LLMModel,
+        token_mgr: token.TokenManager,
+        requester: ProviderAPIRequester,
+    ):
+        self.model_entity = model_entity
+        self.token_mgr = token_mgr
+        self.requester = requester
+
+
+class RuntimeEmbeddingModel:
+    """运行时 Embedding 模型"""
+
+    model_entity: persistence_model.EmbeddingModel
+    """模型数据"""
+
+    token_mgr: token.TokenManager
+    """api key管理器"""
+
+    requester: ProviderAPIRequester
+    """请求器实例"""
+
+    def __init__(
+        self,
+        model_entity: persistence_model.EmbeddingModel,
+        token_mgr: token.TokenManager,
+        requester: ProviderAPIRequester,
+    ):
+        self.model_entity = model_entity
+        self.token_mgr = token_mgr
+        self.requester = requester
+
+
+class ProviderAPIRequester(metaclass=abc.ABCMeta):
+    """Provider API请求器"""
+
+    name: str = None
+
+    ap: app.Application
+
+    default_config: dict[str, typing.Any] = {}
+
+    requester_cfg: dict[str, typing.Any] = {}
+
+    def __init__(self, ap: app.Application, config: dict[str, typing.Any]):
+        self.ap = ap
+        self.requester_cfg = {**self.default_config}
+        self.requester_cfg.update(config)
+
+    async def initialize(self):
+        pass
+
+    @abc.abstractmethod
+    async def invoke_llm(
+        self,
+        query: pipeline_query.Query,
+        model: RuntimeLLMModel,
+        messages: typing.List[provider_message.Message],
+        funcs: typing.List[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.Message:
+        """调用API
+
+        Args:
+            model (RuntimeLLMModel): 使用的模型信息
+            messages (typing.List[llm_entities.Message]): 消息对象列表
+            funcs (typing.List[tools_entities.LLMFunction], optional): 使用的工具函数列表. Defaults to None.
+            extra_args (dict[str, typing.Any], optional): 额外的参数. Defaults to {}.
+            remove_think (bool, optional): 是否移思考中的消息. Defaults to False.
+
+        Returns:
+            llm_entities.Message: 返回消息对象
+        """
+        pass
+
+    async def invoke_llm_stream(
+        self,
+        query: pipeline_query.Query,
+        model: RuntimeLLMModel,
+        messages: typing.List[provider_message.Message],
+        funcs: typing.List[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.MessageChunk:
+        """调用API
+
+        Args:
+            model (RuntimeLLMModel): 使用的模型信息
+            messages (typing.List[provider_message.Message]): 消息对象列表
+            funcs (typing.List[resource_tool.LLMTool], optional): 使用的工具函数列表. Defaults to None.
+            extra_args (dict[str, typing.Any], optional): 额外的参数. Defaults to {}.
+            remove_think (bool, optional): 是否移除思考中的消息. Defaults to False.
+
+        Returns:
+            typing.AsyncGenerator[provider_message.MessageChunk]: 返回消息对象
+        """
+        pass
+
+    async def invoke_embedding(
+        self,
+        model: RuntimeEmbeddingModel,
+        input_text: typing.List[str],
+        extra_args: dict[str, typing.Any] = {},
+    ) -> typing.List[typing.List[float]]:
+        """调用 Embedding API
+
+        Args:
+            model (RuntimeEmbeddingModel): 使用的模型信息
+            input_text (typing.List[str]): 输入文本
+            extra_args (dict[str, typing.Any], optional): 额外的参数. Defaults to {}.
+
+        Returns:
+            typing.List[typing.List[float]]: 返回的 embedding 向量
+        """
+        pass
--- a/src/langbot/pkg/provider/modelmgr/requester.yaml
+++ b/src/langbot/pkg/provider/modelmgr/requester.yaml
@@ -0,0 +1,14 @@
+apiVersion: v1
+kind: ComponentTemplate
+metadata:
+  name: LLMAPIRequester
+  label:
+    en_US: LLM API Requester
+    zh_Hans: LLM API 请求器
+spec:
+  type:
+    - python
+execution:
+  python:
+    path: ./requester.py
+    attr: LLMAPIRequester
--- a/src/langbot/pkg/provider/modelmgr/requesters/302ai.png
+++ b/src/langbot/pkg/provider/modelmgr/requesters/302ai.png
--- a/src/langbot/pkg/provider/modelmgr/requesters/302aichatcmpl.py
+++ b/src/langbot/pkg/provider/modelmgr/requesters/302aichatcmpl.py
@@ -0,0 +1,17 @@
+from __future__ import annotations
+
+import typing
+import openai
+
+from . import chatcmpl
+
+
+class AI302ChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """302.AI ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://api.302.ai/v1',
+        'timeout': 120,
+    }
--- a/src/langbot/pkg/provider/modelmgr/requesters/302aichatcmpl.yaml
+++ b/src/langbot/pkg/provider/modelmgr/requesters/302aichatcmpl.yaml
@@ -0,0 +1,32 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: 302-ai-chat-completions
+  label:
+    en_US: 302.AI
+    zh_Hans: 302.AI
+  icon: 302ai.png
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://api.302.ai/v1
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  - text-embedding
+  provider_category: maas
+execution:
+  python:
+    path: ./302aichatcmpl.py
+    attr: AI302ChatCompletions
--- a/src/langbot/pkg/provider/modelmgr/requesters/init.py
+++ b/src/langbot/pkg/provider/modelmgr/requesters/init.py
--- a/src/langbot/pkg/provider/modelmgr/requesters/anthropic.svg
+++ b/src/langbot/pkg/provider/modelmgr/requesters/anthropic.svg
@@ -0,0 +1,4 @@
+<svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg">
+<rect width="24" height="24" rx="6" fill="#CA9F7B"/>
+<path d="M15.3843 6.43481H12.9687L17.3739 17.5652H19.7896L15.3843 6.43481ZM8.40522 6.43481L4 17.5652H6.4633L7.36417 15.2279H11.9729L12.8737 17.5652H15.337L10.9318 6.43481H8.40522ZM8.16104 13.1607L9.66852 9.24907L11.176 13.1607H8.16104Z" fill="#191918"/>
+</svg>
--- a/src/langbot/pkg/provider/modelmgr/requesters/anthropicmsgs.py
+++ b/src/langbot/pkg/provider/modelmgr/requesters/anthropicmsgs.py
@@ -0,0 +1,370 @@
+from __future__ import annotations
+
+import typing
+import json
+import platform
+import socket
+import anthropic
+import httpx
+
+from .. import errors, requester
+
+from ....utils import image
+import langbot_plugin.api.entities.builtin.resource.tool as resource_tool
+import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
+import langbot_plugin.api.entities.builtin.provider.message as provider_message
+
+
+class AnthropicMessages(requester.ProviderAPIRequester):
+    """Anthropic Messages API 请求器"""
+
+    client: anthropic.AsyncAnthropic
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://api.anthropic.com',
+        'timeout': 120,
+    }
+
+    async def initialize(self):
+        # 兼容 Windows 缺失 TCP_KEEPINTVL 和 TCP_KEEPCNT 的问题
+        if platform.system() == 'Windows':
+            if not hasattr(socket, 'TCP_KEEPINTVL'):
+                socket.TCP_KEEPINTVL = 0
+            if not hasattr(socket, 'TCP_KEEPCNT'):
+                socket.TCP_KEEPCNT = 0
+        httpx_client = anthropic._base_client.AsyncHttpxClientWrapper(
+            base_url=self.requester_cfg['base_url'],
+            # cast to a valid type because mypy doesn't understand our type narrowing
+            timeout=typing.cast(httpx.Timeout, self.requester_cfg['timeout']),
+            limits=anthropic._constants.DEFAULT_CONNECTION_LIMITS,
+            follow_redirects=True,
+            trust_env=True,
+        )
+
+        self.client = anthropic.AsyncAnthropic(
+            api_key='',
+            http_client=httpx_client,
+            base_url=self.requester_cfg['base_url'],
+        )
+
+    async def invoke_llm(
+        self,
+        query: pipeline_query.Query,
+        model: requester.RuntimeLLMModel,
+        messages: typing.List[provider_message.Message],
+        funcs: typing.List[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.Message:
+        self.client.api_key = model.token_mgr.get_token()
+
+        args = extra_args.copy()
+        args['model'] = model.model_entity.name
+
+        # 处理消息
+
+        # system
+        system_role_message = None
+
+        for i, m in enumerate(messages):
+            if m.role == 'system':
+                system_role_message = m
+
+                break
+
+        if system_role_message:
+            messages.pop(i)
+
+        if isinstance(system_role_message, provider_message.Message) and isinstance(system_role_message.content, str):
+            args['system'] = system_role_message.content
+
+        req_messages = []
+
+        for m in messages:
+            if m.role == 'tool':
+                tool_call_id = m.tool_call_id
+
+                req_messages.append(
+                    {
+                        'role': 'user',
+                        'content': [
+                            {
+                                'type': 'tool_result',
+                                'tool_use_id': tool_call_id,
+                                'is_error': False,
+                                'content': [{'type': 'text', 'text': m.content}],
+                            }
+                        ],
+                    }
+                )
+
+                continue
+
+            msg_dict = m.dict(exclude_none=True)
+
+            if isinstance(m.content, str) and m.content.strip() != '':
+                msg_dict['content'] = [{'type': 'text', 'text': m.content}]
+            elif isinstance(m.content, list):
+                for i, ce in enumerate(m.content):
+                    if ce.type == 'image_base64':
+                        image_b64, image_format = await image.extract_b64_and_format(ce.image_base64)
+
+                        alter_image_ele = {
+                            'type': 'image',
+                            'source': {
+                                'type': 'base64',
+                                'media_type': f'image/{image_format}',
+                                'data': image_b64,
+                            },
+                        }
+                        msg_dict['content'][i] = alter_image_ele
+
+            if m.tool_calls:
+                for tool_call in m.tool_calls:
+                    msg_dict['content'].append(
+                        {
+                            'type': 'tool_use',
+                            'id': tool_call.id,
+                            'name': tool_call.function.name,
+                            'input': json.loads(tool_call.function.arguments),
+                        }
+                    )
+
+                del msg_dict['tool_calls']
+
+            req_messages.append(msg_dict)
+
+        args['messages'] = req_messages
+
+        if 'thinking' in args:
+            args['thinking'] = {'type': 'enabled', 'budget_tokens': 10000}
+
+        if funcs:
+            tools = await self.ap.tool_mgr.generate_tools_for_anthropic(funcs)
+
+            if tools:
+                args['tools'] = tools
+
+        try:
+            resp = await self.client.messages.create(**args)
+
+            args = {
+                'content': '',
+                'role': resp.role,
+            }
+            assert type(resp) is anthropic.types.message.Message
+
+            for block in resp.content:
+                if not remove_think and block.type == 'thinking':
+                    args['content'] = '<think>\n' + block.thinking + '\n</think>\n' + args['content']
+                elif block.type == 'text':
+                    args['content'] += block.text
+                elif block.type == 'tool_use':
+                    assert type(block) is anthropic.types.tool_use_block.ToolUseBlock
+                    tool_call = provider_message.ToolCall(
+                        id=block.id,
+                        type='function',
+                        function=provider_message.FunctionCall(name=block.name, arguments=json.dumps(block.input)),
+                    )
+                    if 'tool_calls' not in args:
+                        args['tool_calls'] = []
+                    args['tool_calls'].append(tool_call)
+
+            return provider_message.Message(**args)
+        except anthropic.AuthenticationError as e:
+            raise errors.RequesterError(f'api-key 无效: {e.message}')
+        except anthropic.BadRequestError as e:
+            raise errors.RequesterError(str(e.message))
+        except anthropic.NotFoundError as e:
+            if 'model: ' in str(e):
+                raise errors.RequesterError(f'模型无效: {e.message}')
+            else:
+                raise errors.RequesterError(f'请求地址无效: {e.message}')
+
+    async def invoke_llm_stream(
+        self,
+        query: pipeline_query.Query,
+        model: requester.RuntimeLLMModel,
+        messages: typing.List[provider_message.Message],
+        funcs: typing.List[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.Message:
+        self.client.api_key = model.token_mgr.get_token()
+
+        args = extra_args.copy()
+        args['model'] = model.model_entity.name
+        args['stream'] = True
+
+        # 处理消息
+
+        # system
+        system_role_message = None
+
+        for i, m in enumerate(messages):
+            if m.role == 'system':
+                system_role_message = m
+
+                break
+
+        if system_role_message:
+            messages.pop(i)
+
+        if isinstance(system_role_message, provider_message.Message) and isinstance(system_role_message.content, str):
+            args['system'] = system_role_message.content
+
+        req_messages = []
+
+        for m in messages:
+            if m.role == 'tool':
+                tool_call_id = m.tool_call_id
+
+                req_messages.append(
+                    {
+                        'role': 'user',
+                        'content': [
+                            {
+                                'type': 'tool_result',
+                                'tool_use_id': tool_call_id,
+                                'is_error': False,  # 暂时直接写false
+                                'content': [
+                                    {'type': 'text', 'text': m.content}
+                                ],  # 这里要是list包裹，应该是多个返回的情况？type类型好像也可以填其他的，暂时只写text
+                            }
+                        ],
+                    }
+                )
+
+                continue
+
+            msg_dict = m.dict(exclude_none=True)
+
+            if isinstance(m.content, str) and m.content.strip() != '':
+                msg_dict['content'] = [{'type': 'text', 'text': m.content}]
+            elif isinstance(m.content, list):
+                for i, ce in enumerate(m.content):
+                    if ce.type == 'image_base64':
+                        image_b64, image_format = await image.extract_b64_and_format(ce.image_base64)
+
+                        alter_image_ele = {
+                            'type': 'image',
+                            'source': {
+                                'type': 'base64',
+                                'media_type': f'image/{image_format}',
+                                'data': image_b64,
+                            },
+                        }
+                        msg_dict['content'][i] = alter_image_ele
+            if isinstance(msg_dict['content'], str) and msg_dict['content'] == '':
+                msg_dict['content'] = []  # 这里不知道为什么会莫名有个空导致content为字符
+            if m.tool_calls:
+                for tool_call in m.tool_calls:
+                    msg_dict['content'].append(
+                        {
+                            'type': 'tool_use',
+                            'id': tool_call.id,
+                            'name': tool_call.function.name,
+                            'input': json.loads(tool_call.function.arguments),
+                        }
+                    )
+
+                del msg_dict['tool_calls']
+
+            req_messages.append(msg_dict)
+        if 'thinking' in args:
+            args['thinking'] = {'type': 'enabled', 'budget_tokens': 10000}
+
+        args['messages'] = req_messages
+
+        if funcs:
+            tools = await self.ap.tool_mgr.generate_tools_for_anthropic(funcs)
+
+            if tools:
+                args['tools'] = tools
+
+        try:
+            role = 'assistant'  # 默认角色
+            # chunk_idx = 0
+            think_started = False
+            think_ended = False
+            finish_reason = False
+            content = ''
+            tool_name = ''
+            tool_id = ''
+            async for chunk in await self.client.messages.create(**args):
+                tool_call = {'id': None, 'function': {'name': None, 'arguments': None}, 'type': 'function'}
+                if isinstance(
+                    chunk, anthropic.types.raw_content_block_start_event.RawContentBlockStartEvent
+                ):  # 记录开始
+                    if chunk.content_block.type == 'tool_use':
+                        if chunk.content_block.name is not None:
+                            tool_name = chunk.content_block.name
+                        if chunk.content_block.id is not None:
+                            tool_id = chunk.content_block.id
+
+                        tool_call['function']['name'] = tool_name
+                        tool_call['function']['arguments'] = ''
+                        tool_call['id'] = tool_id
+
+                    if not remove_think:
+                        if chunk.content_block.type == 'thinking' and not remove_think:
+                            think_started = True
+                        elif chunk.content_block.type == 'text' and chunk.index != 0 and not remove_think:
+                            think_ended = True
+                        continue
+                elif isinstance(chunk, anthropic.types.raw_content_block_delta_event.RawContentBlockDeltaEvent):
+                    if chunk.delta.type == 'thinking_delta':
+                        if think_started:
+                            think_started = False
+                            content = '<think>\n' + chunk.delta.thinking
+                        elif remove_think:
+                            continue
+                        else:
+                            content = chunk.delta.thinking
+                    elif chunk.delta.type == 'text_delta':
+                        if think_ended:
+                            think_ended = False
+                            content = '\n</think>\n' + chunk.delta.text
+                        else:
+                            content = chunk.delta.text
+                    elif chunk.delta.type == 'input_json_delta':
+                        tool_call['function']['arguments'] = chunk.delta.partial_json
+                        tool_call['function']['name'] = tool_name
+                        tool_call['id'] = tool_id
+                elif isinstance(chunk, anthropic.types.raw_content_block_stop_event.RawContentBlockStopEvent):
+                    continue  # 记录raw_content_block结束的
+
+                elif isinstance(chunk, anthropic.types.raw_message_delta_event.RawMessageDeltaEvent):
+                    if chunk.delta.stop_reason == 'end_turn':
+                        finish_reason = True
+                elif isinstance(chunk, anthropic.types.raw_message_stop_event.RawMessageStopEvent):
+                    continue  # 这个好像是完全结束
+                else:
+                    # print(chunk)
+                    self.ap.logger.debug(f'anthropic chunk: {chunk}')
+                    continue
+
+                args = {
+                    'content': content,
+                    'role': role,
+                    'is_final': finish_reason,
+                    'tool_calls': None if tool_call['id'] is None else [tool_call],
+                }
+                # if chunk_idx == 0:
+                #     chunk_idx += 1
+                #     continue
+
+                # assert type(chunk) is anthropic.types.message.Chunk
+
+                yield provider_message.MessageChunk(**args)
+
+            # return llm_entities.Message(**args)
+        except anthropic.AuthenticationError as e:
+            raise errors.RequesterError(f'api-key 无效: {e.message}')
+        except anthropic.BadRequestError as e:
+            raise errors.RequesterError(str(e.message))
+        except anthropic.NotFoundError as e:
+            if 'model: ' in str(e):
+                raise errors.RequesterError(f'模型无效: {e.message}')
+            else:
+                raise errors.RequesterError(f'请求地址无效: {e.message}')
--- a/src/langbot/pkg/provider/modelmgr/requesters/anthropicmsgs.yaml
+++ b/src/langbot/pkg/provider/modelmgr/requesters/anthropicmsgs.yaml
@@ -0,0 +1,31 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: anthropic-messages
+  label:
+    en_US: Anthropic
+    zh_Hans: Anthropic
+  icon: anthropic.svg
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://api.anthropic.com
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  provider_category: manufacturer
+execution:
+  python:
+    path: ./anthropicmsgs.py
+    attr: AnthropicMessages
--- a/src/langbot/pkg/provider/modelmgr/requesters/bailian.png
+++ b/src/langbot/pkg/provider/modelmgr/requesters/bailian.png
--- a/src/langbot/pkg/provider/modelmgr/requesters/bailianchatcmpl.py
+++ b/src/langbot/pkg/provider/modelmgr/requesters/bailianchatcmpl.py
@@ -0,0 +1,242 @@
+from __future__ import annotations
+
+import typing
+import dashscope
+import openai
+
+from . import modelscopechatcmpl
+from .. import requester
+import langbot_plugin.api.entities.builtin.resource.tool as resource_tool
+import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
+import langbot_plugin.api.entities.builtin.provider.message as provider_message
+
+
+class BailianChatCompletions(modelscopechatcmpl.ModelScopeChatCompletions):
+    """阿里云百炼大模型平台 ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://dashscope.aliyuncs.com/compatible-mode/v1',
+        'timeout': 120,
+    }
+
+    async def _closure_stream(
+        self,
+        query: pipeline_query.Query,
+        req_messages: list[dict],
+        use_model: requester.RuntimeLLMModel,
+        use_funcs: list[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.Message | typing.AsyncGenerator[provider_message.MessageChunk, None]:
+        self.client.api_key = use_model.token_mgr.get_token()
+
+        args = {}
+        args['model'] = use_model.model_entity.name
+
+        if use_funcs:
+            tools = await self.ap.tool_mgr.generate_tools_for_openai(use_funcs)
+
+            if tools:
+                args['tools'] = tools
+
+        # 设置此次请求中的messages
+        messages = req_messages.copy()
+
+        is_use_dashscope_call = False  # 是否使用阿里原生库调用
+        is_enable_multi_model = True  # 是否支持多轮对话
+        use_time_num = 0  # 模型已调用次数，防止存在多文件时重复调用
+        use_time_ids = []  # 已调用的ID列表
+        message_id = 0  # 记录消息序号
+
+        for msg in messages:
+            # print(msg)
+            if 'content' in msg and isinstance(msg['content'], list):
+                for me in msg['content']:
+                    if me['type'] == 'image_base64':
+                        me['image_url'] = {'url': me['image_base64']}
+                        me['type'] = 'image_url'
+                        del me['image_base64']
+                    elif me['type'] == 'file_url' and '.' in me.get('file_name', ''):
+                        # 1. 视频文件推理
+                        # https://bailian.console.aliyun.com/?tab=doc#/doc/?type=model&url=2845871
+                        file_type = me.get('file_name').lower().split('.')[-1]
+                        if file_type in ['mp4', 'avi', 'mkv', 'mov', 'flv', 'wmv']:
+                            me['type'] = 'video_url'
+                            me['video_url'] = {'url': me['file_url']}
+                            del me['file_url']
+                            del me['file_name']
+                            use_time_num += 1
+                            use_time_ids.append(message_id)
+                            is_enable_multi_model = False
+                        # 2. 语音文件识别, 无法通过openai的audio字段传递，暂时不支持
+                        # https://bailian.console.aliyun.com/?tab=doc#/doc/?type=model&url=2979031
+                        elif file_type in [
+                            'aac',
+                            'amr',
+                            'aiff',
+                            'flac',
+                            'm4a',
+                            'mp3',
+                            'mpeg',
+                            'ogg',
+                            'opus',
+                            'wav',
+                            'webm',
+                            'wma',
+                        ]:
+                            me['audio'] = me['file_url']
+                            me['type'] = 'audio'
+                            del me['file_url']
+                            del me['type']
+                            del me['file_name']
+                            is_use_dashscope_call = True
+                            use_time_num += 1
+                            use_time_ids.append(message_id)
+                            is_enable_multi_model = False
+            message_id += 1
+
+        # 使用列表推导式，保留不在 use_time_ids[:-1] 中的元素，仅保留最后一个多媒体消息
+        if not is_enable_multi_model and use_time_num > 1:
+            messages = [msg for idx, msg in enumerate(messages) if idx not in use_time_ids[:-1]]
+
+        if not is_enable_multi_model:
+            messages = [msg for msg in messages if 'resp_message_id' not in msg]
+
+        args['messages'] = messages
+        args['stream'] = True
+
+        # 流式处理状态
+        # tool_calls_map: dict[str, provider_message.ToolCall] = {}
+        chunk_idx = 0
+        thinking_started = False
+        thinking_ended = False
+        role = 'assistant'  # 默认角色
+
+        if is_use_dashscope_call:
+            response = dashscope.MultiModalConversation.call(
+                # 若没有配置环境变量，请用百炼API Key将下行替换为：api_key = "sk-xxx"
+                api_key=use_model.token_mgr.get_token(),
+                model=use_model.model_entity.name,
+                messages=messages,
+                result_format='message',
+                asr_options={
+                    # "language": "zh", # 可选，若已知音频的语种，可通过该参数指定待识别语种，以提升识别准确率
+                    'enable_lid': True,
+                    'enable_itn': False,
+                },
+                stream=True,
+            )
+            content_length_list = []
+            previous_length = 0  # 记录上一次的内容长度
+            for res in response:
+                chunk = res['output']
+                # 解析 chunk 数据
+                if hasattr(chunk, 'choices') and chunk.choices:
+                    choice = chunk.choices[0]
+                    delta_content = choice['message'].content[0]['text']
+                    finish_reason = choice['finish_reason']
+                    content_length_list.append(len(delta_content))
+                else:
+                    delta_content = ''
+                    finish_reason = None
+
+                # 跳过空的第一个 chunk（只有 role 没有内容）
+                if chunk_idx == 0 and not delta_content:
+                    chunk_idx += 1
+                    continue
+
+                # 检查 content_length_list 是否有足够的数据
+                if len(content_length_list) >= 2:
+                    now_content = delta_content[previous_length : content_length_list[-1]]
+                    previous_length = content_length_list[-1]  # 更新上一次的长度
+                else:
+                    now_content = delta_content  # 第一次循环时直接使用 delta_content
+                    previous_length = len(delta_content)  # 更新上一次的长度
+
+                # 构建 MessageChunk - 只包含增量内容
+                chunk_data = {
+                    'role': role,
+                    'content': now_content if now_content else None,
+                    'is_final': bool(finish_reason) and finish_reason != 'null',
+                }
+
+                # 移除 None 值
+                chunk_data = {k: v for k, v in chunk_data.items() if v is not None}
+                yield provider_message.MessageChunk(**chunk_data)
+                chunk_idx += 1
+        else:
+            async for chunk in self._req_stream(args, extra_body=extra_args):
+                # 解析 chunk 数据
+                if hasattr(chunk, 'choices') and chunk.choices:
+                    choice = chunk.choices[0]
+                    delta = choice.delta.model_dump() if hasattr(choice, 'delta') else {}
+                    finish_reason = getattr(choice, 'finish_reason', None)
+                else:
+                    delta = {}
+                    finish_reason = None
+
+                # 从第一个 chunk 获取 role，后续使用这个 role
+                if 'role' in delta and delta['role']:
+                    role = delta['role']
+
+                # 获取增量内容
+                delta_content = delta.get('content', '')
+                reasoning_content = delta.get('reasoning_content', '')
+
+                # 处理 reasoning_content
+                if reasoning_content:
+                    # accumulated_reasoning += reasoning_content
+                    # 如果设置了 remove_think，跳过 reasoning_content
+                    if remove_think:
+                        chunk_idx += 1
+                        continue
+
+                    # 第一次出现 reasoning_content，添加 <think> 开始标签
+                    if not thinking_started:
+                        thinking_started = True
+                        delta_content = '<think>\n' + reasoning_content
+                    else:
+                        # 继续输出 reasoning_content
+                        delta_content = reasoning_content
+                elif thinking_started and not thinking_ended and delta_content:
+                    # reasoning_content 结束，normal content 开始，添加 </think> 结束标签
+                    thinking_ended = True
+                    delta_content = '\n</think>\n' + delta_content
+
+                # 处理工具调用增量
+                if delta.get('tool_calls'):
+                    for tool_call in delta['tool_calls']:
+                        if tool_call['id'] != '':
+                            tool_id = tool_call['id']
+                        if tool_call['function']['name'] is not None:
+                            tool_name = tool_call['function']['name']
+
+                        if tool_call['type'] is None:
+                            tool_call['type'] = 'function'
+                        tool_call['id'] = tool_id
+                        tool_call['function']['name'] = tool_name
+                        tool_call['function']['arguments'] = (
+                            '' if tool_call['function']['arguments'] is None else tool_call['function']['arguments']
+                        )
+
+                # 跳过空的第一个 chunk（只有 role 没有内容）
+                if chunk_idx == 0 and not delta_content and not reasoning_content and not delta.get('tool_calls'):
+                    chunk_idx += 1
+                    continue
+
+                # 构建 MessageChunk - 只包含增量内容
+                chunk_data = {
+                    'role': role,
+                    'content': delta_content if delta_content else None,
+                    'tool_calls': delta.get('tool_calls'),
+                    'is_final': bool(finish_reason),
+                }
+
+                # 移除 None 值
+                chunk_data = {k: v for k, v in chunk_data.items() if v is not None}
+
+                yield provider_message.MessageChunk(**chunk_data)
+                chunk_idx += 1
+                # return
--- a/src/langbot/pkg/provider/modelmgr/requesters/bailianchatcmpl.yaml
+++ b/src/langbot/pkg/provider/modelmgr/requesters/bailianchatcmpl.yaml
@@ -0,0 +1,31 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: bailian-chat-completions
+  label:
+    en_US: Aliyun Bailian
+    zh_Hans: 阿里云百炼
+  icon: bailian.png
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://dashscope.aliyuncs.com/compatible-mode/v1
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  provider_category: maas
+execution:
+  python:
+    path: ./bailianchatcmpl.py
+    attr: BailianChatCompletions
--- a/src/langbot/pkg/provider/modelmgr/requesters/chatcmpl.py
+++ b/src/langbot/pkg/provider/modelmgr/requesters/chatcmpl.py
@@ -0,0 +1,406 @@
+from __future__ import annotations
+
+import asyncio
+import typing
+
+import openai
+import openai.types.chat.chat_completion as chat_completion
+import httpx
+
+from .. import errors, requester
+import langbot_plugin.api.entities.builtin.resource.tool as resource_tool
+import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
+import langbot_plugin.api.entities.builtin.provider.message as provider_message
+
+
+class OpenAIChatCompletions(requester.ProviderAPIRequester):
+    """OpenAI ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://api.openai.com/v1',
+        'timeout': 120,
+    }
+
+    async def initialize(self):
+        self.client = openai.AsyncClient(
+            api_key='',
+            base_url=self.requester_cfg['base_url'].replace(' ', ''),
+            timeout=self.requester_cfg['timeout'],
+            http_client=httpx.AsyncClient(trust_env=True, timeout=self.requester_cfg['timeout']),
+        )
+
+    async def _req(
+        self,
+        args: dict,
+        extra_body: dict = {},
+    ) -> chat_completion.ChatCompletion:
+        return await self.client.chat.completions.create(**args, extra_body=extra_body)
+
+    async def _req_stream(
+        self,
+        args: dict,
+        extra_body: dict = {},
+    ):
+        async for chunk in await self.client.chat.completions.create(**args, extra_body=extra_body):
+            yield chunk
+
+    async def _make_msg(
+        self,
+        chat_completion: chat_completion.ChatCompletion,
+        remove_think: bool = False,
+    ) -> provider_message.Message:
+        chatcmpl_message = chat_completion.choices[0].message.model_dump()
+
+        # 确保 role 字段存在且不为 None
+        if 'role' not in chatcmpl_message or chatcmpl_message['role'] is None:
+            chatcmpl_message['role'] = 'assistant'
+
+        # 处理思维链
+        content = chatcmpl_message.get('content', '')
+        reasoning_content = chatcmpl_message.get('reasoning_content', None)
+
+        processed_content, _ = await self._process_thinking_content(
+            content=content, reasoning_content=reasoning_content, remove_think=remove_think
+        )
+
+        chatcmpl_message['content'] = processed_content
+
+        # 移除 reasoning_content 字段，避免传递给 Message
+        if 'reasoning_content' in chatcmpl_message:
+            del chatcmpl_message['reasoning_content']
+
+        message = provider_message.Message(**chatcmpl_message)
+
+        return message
+
+    async def _process_thinking_content(
+        self,
+        content: str,
+        reasoning_content: str = None,
+        remove_think: bool = False,
+    ) -> tuple[str, str]:
+        """处理思维链内容
+
+        Args:
+            content: 原始内容
+            reasoning_content: reasoning_content 字段内容
+            remove_think: 是否移除思维链
+
+        Returns:
+            (处理后的内容, 提取的思维链内容)
+        """
+        thinking_content = ''
+
+        # 1. 从 reasoning_content 提取思维链
+        if reasoning_content:
+            thinking_content = reasoning_content
+
+        # 2. 从 content 中提取 <think> 标签内容
+        if content and '<think>' in content and '</think>' in content:
+            import re
+
+            think_pattern = r'<think>(.*?)</think>'
+            think_matches = re.findall(think_pattern, content, re.DOTALL)
+            if think_matches:
+                # 如果已有 reasoning_content，则追加
+                if thinking_content:
+                    thinking_content += '\n' + '\n'.join(think_matches)
+                else:
+                    thinking_content = '\n'.join(think_matches)
+                # 移除 content 中的 <think> 标签
+                content = re.sub(think_pattern, '', content, flags=re.DOTALL).strip()
+
+        # 3. 根据 remove_think 参数决定是否保留思维链
+        if remove_think:
+            return content, ''
+        else:
+            # 如果有思维链内容，将其以 <think> 格式添加到 content 开头
+            if thinking_content:
+                content = f'<think>\n{thinking_content}\n</think>\n{content}'.strip()
+            return content, thinking_content
+
+    async def _closure_stream(
+        self,
+        query: pipeline_query.Query,
+        req_messages: list[dict],
+        use_model: requester.RuntimeLLMModel,
+        use_funcs: list[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.MessageChunk:
+        self.client.api_key = use_model.token_mgr.get_token()
+
+        args = {}
+        args['model'] = use_model.model_entity.name
+
+        if use_funcs:
+            tools = await self.ap.tool_mgr.generate_tools_for_openai(use_funcs)
+            if tools:
+                args['tools'] = tools
+
+        # 设置此次请求中的messages
+        messages = req_messages.copy()
+
+        # 检查vision
+        for msg in messages:
+            if 'content' in msg and isinstance(msg['content'], list):
+                for me in msg['content']:
+                    if me['type'] == 'image_base64':
+                        me['image_url'] = {'url': me['image_base64']}
+                        me['type'] = 'image_url'
+                        del me['image_base64']
+
+        args['messages'] = messages
+        args['stream'] = True
+
+        # 流式处理状态
+        # tool_calls_map: dict[str, provider_message.ToolCall] = {}
+        chunk_idx = 0
+        thinking_started = False
+        thinking_ended = False
+        role = 'assistant'  # 默认角色
+        tool_id = ''
+        tool_name = ''
+        # accumulated_reasoning = ''  # 仅用于判断何时结束思维链
+
+        async for chunk in self._req_stream(args, extra_body=extra_args):
+            # 解析 chunk 数据
+
+            if hasattr(chunk, 'choices') and chunk.choices:
+                choice = chunk.choices[0]
+                delta = choice.delta.model_dump() if hasattr(choice, 'delta') else {}
+
+                finish_reason = getattr(choice, 'finish_reason', None)
+            else:
+                delta = {}
+                finish_reason = None
+            # 从第一个 chunk 获取 role，后续使用这个 role
+            if 'role' in delta and delta['role']:
+                role = delta['role']
+
+            # 获取增量内容
+            delta_content = delta.get('content', '')
+            reasoning_content = delta.get('reasoning_content', '')
+
+            # 处理 reasoning_content
+            if reasoning_content:
+                # accumulated_reasoning += reasoning_content
+                # 如果设置了 remove_think，跳过 reasoning_content
+                if remove_think:
+                    chunk_idx += 1
+                    continue
+
+                # 第一次出现 reasoning_content，添加 <think> 开始标签
+                if not thinking_started:
+                    thinking_started = True
+                    delta_content = '<think>\n' + reasoning_content
+                else:
+                    # 继续输出 reasoning_content
+                    delta_content = reasoning_content
+            elif thinking_started and not thinking_ended and delta_content:
+                # reasoning_content 结束，normal content 开始，添加 </think> 结束标签
+                thinking_ended = True
+                delta_content = '\n</think>\n' + delta_content
+
+            # 处理 content 中已有的 <think> 标签（如果需要移除）
+            # if delta_content and remove_think and '<think>' in delta_content:
+            #     import re
+            #
+            #     # 移除 <think> 标签及其内容
+            #     delta_content = re.sub(r'<think>.*?</think>', '', delta_content, flags=re.DOTALL)
+
+            # 处理工具调用增量
+            # delta_tool_calls = None
+            if delta.get('tool_calls'):
+                for tool_call in delta['tool_calls']:
+                    if tool_call['id'] and tool_call['function']['name']:
+                        tool_id = tool_call['id']
+                        tool_name = tool_call['function']['name']
+                    else:
+                        tool_call['id'] = tool_id
+                        tool_call['function']['name'] = tool_name
+                    if tool_call['type'] is None:
+                        tool_call['type'] = 'function'
+
+            # 跳过空的第一个 chunk（只有 role 没有内容）
+            if chunk_idx == 0 and not delta_content and not reasoning_content and not delta.get('tool_calls'):
+                chunk_idx += 1
+                continue
+            # 构建 MessageChunk - 只包含增量内容
+            chunk_data = {
+                'role': role,
+                'content': delta_content if delta_content else None,
+                'tool_calls': delta.get('tool_calls'),
+                'is_final': bool(finish_reason),
+            }
+
+            # 移除 None 值
+            chunk_data = {k: v for k, v in chunk_data.items() if v is not None}
+
+            yield provider_message.MessageChunk(**chunk_data)
+            chunk_idx += 1
+
+    async def _closure(
+        self,
+        query: pipeline_query.Query,
+        req_messages: list[dict],
+        use_model: requester.RuntimeLLMModel,
+        use_funcs: list[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.Message:
+        self.client.api_key = use_model.token_mgr.get_token()
+
+        args = {}
+        args['model'] = use_model.model_entity.name
+
+        if use_funcs:
+            tools = await self.ap.tool_mgr.generate_tools_for_openai(use_funcs)
+
+            if tools:
+                args['tools'] = tools
+
+        # 设置此次请求中的messages
+        messages = req_messages.copy()
+
+        # 检查vision
+        for msg in messages:
+            if 'content' in msg and isinstance(msg['content'], list):
+                for me in msg['content']:
+                    if me['type'] == 'image_base64':
+                        me['image_url'] = {'url': me['image_base64']}
+                        me['type'] = 'image_url'
+                        del me['image_base64']
+
+        args['messages'] = messages
+
+        # 发送请求
+
+        resp = await self._req(args, extra_body=extra_args)
+        # 处理请求结果
+        message = await self._make_msg(resp, remove_think)
+
+        return message
+
+    async def invoke_llm(
+        self,
+        query: pipeline_query.Query,
+        model: requester.RuntimeLLMModel,
+        messages: typing.List[provider_message.Message],
+        funcs: typing.List[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.Message:
+        req_messages = []  # req_messages 仅用于类内，外部同步由 query.messages 进行
+        for m in messages:
+            msg_dict = m.dict(exclude_none=True)
+            content = msg_dict.get('content')
+            if isinstance(content, list):
+                # 检查 content 列表中是否每个部分都是文本
+                if all(isinstance(part, dict) and part.get('type') == 'text' for part in content):
+                    # 将所有文本部分合并为一个字符串
+                    msg_dict['content'] = '\n'.join(part['text'] for part in content)
+            req_messages.append(msg_dict)
+
+        try:
+            msg = await self._closure(
+                query=query,
+                req_messages=req_messages,
+                use_model=model,
+                use_funcs=funcs,
+                extra_args=extra_args,
+                remove_think=remove_think,
+            )
+            return msg
+        except asyncio.TimeoutError:
+            raise errors.RequesterError('请求超时')
+        except openai.BadRequestError as e:
+            if 'context_length_exceeded' in e.message:
+                raise errors.RequesterError(f'上文过长，请重置会话: {e.message}')
+            else:
+                raise errors.RequesterError(f'请求参数错误: {e.message}')
+        except openai.AuthenticationError as e:
+            raise errors.RequesterError(f'无效的 api-key: {e.message}')
+        except openai.NotFoundError as e:
+            raise errors.RequesterError(f'请求路径错误: {e.message}')
+        except openai.RateLimitError as e:
+            raise errors.RequesterError(f'请求过于频繁或余额不足: {e.message}')
+        except openai.APIError as e:
+            raise errors.RequesterError(f'请求错误: {e.message}')
+
+    async def invoke_embedding(
+        self,
+        model: requester.RuntimeEmbeddingModel,
+        input_text: list[str],
+        extra_args: dict[str, typing.Any] = {},
+    ) -> list[list[float]]:
+        """调用 Embedding API"""
+        self.client.api_key = model.token_mgr.get_token()
+
+        args = {
+            'model': model.model_entity.name,
+            'input': input_text,
+        }
+
+        if model.model_entity.extra_args:
+            args.update(model.model_entity.extra_args)
+
+        args.update(extra_args)
+
+        try:
+            resp = await self.client.embeddings.create(**args)
+
+            return [d.embedding for d in resp.data]
+        except asyncio.TimeoutError:
+            raise errors.RequesterError('请求超时')
+        except openai.BadRequestError as e:
+            raise errors.RequesterError(f'请求参数错误: {e.message}')
+
+    async def invoke_llm_stream(
+        self,
+        query: pipeline_query.Query,
+        model: requester.RuntimeLLMModel,
+        messages: typing.List[provider_message.Message],
+        funcs: typing.List[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.MessageChunk:
+        req_messages = []  # req_messages 仅用于类内，外部同步由 query.messages 进行
+        for m in messages:
+            msg_dict = m.dict(exclude_none=True)
+            content = msg_dict.get('content')
+            if isinstance(content, list):
+                # 检查 content 列表中是否每个部分都是文本
+                if all(isinstance(part, dict) and part.get('type') == 'text' for part in content):
+                    # 将所有文本部分合并为一个字符串
+                    msg_dict['content'] = '\n'.join(part['text'] for part in content)
+            req_messages.append(msg_dict)
+
+        try:
+            async for item in self._closure_stream(
+                query=query,
+                req_messages=req_messages,
+                use_model=model,
+                use_funcs=funcs,
+                extra_args=extra_args,
+                remove_think=remove_think,
+            ):
+                yield item
+
+        except asyncio.TimeoutError:
+            raise errors.RequesterError('请求超时')
+        except openai.BadRequestError as e:
+            if 'context_length_exceeded' in e.message:
+                raise errors.RequesterError(f'上文过长，请重置会话: {e.message}')
+            else:
+                raise errors.RequesterError(f'请求参数错误: {e.message}')
+        except openai.AuthenticationError as e:
+            raise errors.RequesterError(f'无效的 api-key: {e.message}')
+        except openai.NotFoundError as e:
+            raise errors.RequesterError(f'请求路径错误: {e.message}')
+        except openai.RateLimitError as e:
+            raise errors.RequesterError(f'请求过于频繁或余额不足: {e.message}')
+        except openai.APIError as e:
+            raise errors.RequesterError(f'请求错误: {e.message}')
--- a/src/langbot/pkg/provider/modelmgr/requesters/chatcmpl.yaml
+++ b/src/langbot/pkg/provider/modelmgr/requesters/chatcmpl.yaml
@@ -0,0 +1,32 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: openai-chat-completions
+  label:
+    en_US: OpenAI
+    zh_Hans: OpenAI
+  icon: openai.svg
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://api.openai.com/v1
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  - text-embedding
+  provider_category: manufacturer
+execution:
+  python:
+    path: ./chatcmpl.py
+    attr: OpenAIChatCompletions
--- a/src/langbot/pkg/provider/modelmgr/requesters/compshare.png
+++ b/src/langbot/pkg/provider/modelmgr/requesters/compshare.png
--- a/src/langbot/pkg/provider/modelmgr/requesters/compsharechatcmpl.py
+++ b/src/langbot/pkg/provider/modelmgr/requesters/compsharechatcmpl.py
@@ -0,0 +1,17 @@
+from __future__ import annotations
+
+import typing
+import openai
+
+from . import chatcmpl
+
+
+class CompShareChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """CompShare ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://api.modelverse.cn/v1',
+        'timeout': 120,
+    }
--- a/src/langbot/pkg/provider/modelmgr/requesters/compsharechatcmpl.yaml
+++ b/src/langbot/pkg/provider/modelmgr/requesters/compsharechatcmpl.yaml
@@ -0,0 +1,31 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: compshare-chat-completions
+  label:
+    en_US: CompShare
+    zh_Hans: 优云智算
+  icon: compshare.png
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://api.modelverse.cn/v1
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  provider_category: maas
+execution:
+  python:
+    path: ./compsharechatcmpl.py
+    attr: CompShareChatCompletions
--- a/src/langbot/pkg/provider/modelmgr/requesters/deepseek.svg
+++ b/src/langbot/pkg/provider/modelmgr/requesters/deepseek.svg
@@ -0,0 +1,3 @@
+<svg width="60" height="50" viewBox="0 0 60 50" version="1.1" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" xml:space="preserve" xmlns:serif="http://www.serif.com/" style="fill-rule:evenodd;clip-rule:evenodd;stroke-linejoin:round;stroke-miterlimit:2;">
+    <path id="path" d="M55.613,3.471C55.018,3.179 54.761,3.735 54.413,4.018C54.294,4.109 54.193,4.227 54.092,4.337C53.222,5.266 52.206,5.876 50.878,5.803C48.936,5.694 47.278,6.305 45.813,7.79C45.501,5.959 44.466,4.865 42.891,4.164C42.067,3.799 41.233,3.435 40.656,2.642C40.253,2.077 40.143,1.448 39.942,0.829C39.814,0.455 39.685,0.073 39.255,0.009C38.788,-0.064 38.605,0.328 38.421,0.656C37.689,1.995 37.405,3.471 37.432,4.965C37.496,8.327 38.916,11.006 41.737,12.91C42.058,13.129 42.14,13.347 42.039,13.666C41.847,14.323 41.618,14.96 41.416,15.616C41.288,16.035 41.096,16.127 40.647,15.944C39.099,15.297 37.762,14.341 36.58,13.184C34.575,11.243 32.761,9.102 30.499,7.425C29.968,7.033 29.436,6.669 28.887,6.323C26.579,4.082 29.189,2.241 29.794,2.022C30.425,1.795 30.013,1.011 27.971,1.02C25.928,1.029 24.06,1.713 21.679,2.624C21.331,2.761 20.964,2.861 20.589,2.943C18.427,2.533 16.183,2.441 13.838,2.706C9.424,3.198 5.898,5.284 3.306,8.847C0.191,13.129 -0.541,17.994 0.356,23.069C1.3,28.417 4.029,32.845 8.224,36.308C12.575,39.897 17.584,41.656 23.3,41.319C26.771,41.118 30.636,40.654 34.996,36.964C36.095,37.51 37.249,37.729 39.163,37.893C40.638,38.03 42.058,37.82 43.157,37.592C44.878,37.228 44.759,35.633 44.137,35.342C39.09,32.991 40.198,33.948 39.191,33.173C41.755,30.139 45.62,26.987 47.132,16.773C47.251,15.962 47.15,15.452 47.132,14.796C47.122,14.395 47.214,14.241 47.672,14.195C48.936,14.049 50.163,13.703 51.29,13.083C54.56,11.298 55.878,8.364 56.19,4.847C56.236,4.309 56.181,3.753 55.613,3.471ZM27.119,35.123C22.228,31.278 19.856,30.012 18.876,30.066C17.96,30.121 18.125,31.169 18.326,31.852C18.537,32.526 18.812,32.991 19.196,33.583C19.462,33.975 19.645,34.558 18.931,34.996C17.356,35.971 14.617,34.667 14.489,34.604C11.301,32.727 8.636,30.249 6.758,26.859C4.945,23.597 3.892,20.099 3.718,16.363C3.672,15.461 3.938,15.142 4.835,14.979C6.017,14.76 7.235,14.714 8.416,14.887C13.408,15.616 17.658,17.848 21.221,21.384C23.254,23.397 24.793,25.802 26.377,28.153C28.062,30.649 29.876,33.028 32.184,34.977C32.999,35.661 33.649,36.18 34.272,36.563C32.395,36.772 29.262,36.818 27.119,35.123ZM29.464,20.044C29.464,19.643 29.784,19.325 30.187,19.325C30.279,19.325 30.361,19.343 30.435,19.37C30.535,19.407 30.627,19.461 30.7,19.543C30.828,19.671 30.902,19.853 30.902,20.044C30.902,20.445 30.581,20.764 30.178,20.764C29.775,20.764 29.464,20.445 29.464,20.044ZM36.745,23.78C36.278,23.971 35.811,24.135 35.362,24.153C34.666,24.19 33.906,23.907 33.494,23.561C32.853,23.023 32.395,22.723 32.202,21.784C32.12,21.384 32.166,20.764 32.239,20.409C32.404,19.643 32.221,19.151 31.68,18.705C31.241,18.34 30.682,18.24 30.068,18.24C29.839,18.24 29.629,18.14 29.473,18.058C29.216,17.93 29.006,17.612 29.207,17.22C29.271,17.092 29.583,16.783 29.656,16.728C30.49,16.254 31.451,16.409 32.34,16.764C33.164,17.101 33.787,17.721 34.684,18.596C35.6,19.652 35.765,19.944 36.287,20.737C36.7,21.356 37.075,21.994 37.331,22.723C37.487,23.179 37.286,23.552 36.745,23.78Z" style="fill:rgb(77,107,254);fill-rule:nonzero;"/>
+</svg>
--- a/src/langbot/pkg/provider/modelmgr/requesters/deepseekchatcmpl.py
+++ b/src/langbot/pkg/provider/modelmgr/requesters/deepseekchatcmpl.py
@@ -0,0 +1,60 @@
+from __future__ import annotations
+
+import typing
+
+from . import chatcmpl
+from .. import errors, requester
+import langbot_plugin.api.entities.builtin.resource.tool as resource_tool
+import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
+import langbot_plugin.api.entities.builtin.provider.message as provider_message
+
+
+class DeepseekChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """Deepseek ChatCompletion API 请求器"""
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://api.deepseek.com',
+        'timeout': 120,
+    }
+
+    async def _closure(
+        self,
+        query: pipeline_query.Query,
+        req_messages: list[dict],
+        use_model: requester.RuntimeLLMModel,
+        use_funcs: list[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.Message:
+        self.client.api_key = use_model.token_mgr.get_token()
+
+        args = {}
+        args['model'] = use_model.model_entity.name
+
+        if use_funcs:
+            tools = await self.ap.tool_mgr.generate_tools_for_openai(use_funcs)
+
+            if tools:
+                args['tools'] = tools
+
+        # 设置此次请求中的messages
+        messages = req_messages
+
+        # deepseek 不支持多模态，把content都转换成纯文字
+        for m in messages:
+            if 'content' in m and isinstance(m['content'], list):
+                m['content'] = ' '.join([c['text'] for c in m['content']])
+
+        args['messages'] = messages
+
+        # 发送请求
+        resp = await self._req(args, extra_body=extra_args)
+
+        # print(resp)
+
+        if resp is None:
+            raise errors.RequesterError('接口返回为空，请确定模型提供商服务是否正常')
+        # 处理请求结果
+        message = await self._make_msg(resp, remove_think)
+
+        return message
--- a/src/langbot/pkg/provider/modelmgr/requesters/deepseekchatcmpl.yaml
+++ b/src/langbot/pkg/provider/modelmgr/requesters/deepseekchatcmpl.yaml
@@ -0,0 +1,31 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: deepseek-chat-completions
+  label:
+    en_US: DeepSeek
+    zh_Hans: DeepSeek
+  icon: deepseek.svg
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://api.deepseek.com
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  provider_category: manufacturer
+execution:
+  python:
+    path: ./deepseekchatcmpl.py
+    attr: DeepseekChatCompletions
--- a/src/langbot/pkg/provider/modelmgr/requesters/gemini.svg
+++ b/src/langbot/pkg/provider/modelmgr/requesters/gemini.svg
@@ -0,0 +1 @@
+<svg height="1em" style="flex:none;line-height:1" viewBox="0 0 24 24" width="1em" xmlns="http://www.w3.org/2000/svg"><title>Gemini</title><defs><linearGradient id="lobe-icons-gemini-fill" x1="0%" x2="68.73%" y1="100%" y2="30.395%"><stop offset="0%" stop-color="#1C7DFF"></stop><stop offset="52.021%" stop-color="#1C69FF"></stop><stop offset="100%" stop-color="#F0DCD6"></stop></linearGradient></defs><path d="M12 24A14.304 14.304 0 000 12 14.304 14.304 0 0012 0a14.305 14.305 0 0012 12 14.305 14.305 0 00-12 12" fill="url(#lobe-icons-gemini-fill)" fill-rule="nonzero"></path></svg>
--- a/src/langbot/pkg/provider/modelmgr/requesters/geminichatcmpl.py
+++ b/src/langbot/pkg/provider/modelmgr/requesters/geminichatcmpl.py
@@ -0,0 +1,142 @@
+from __future__ import annotations
+
+import typing
+
+from . import chatcmpl
+
+import uuid
+
+from .. import requester
+import langbot_plugin.api.entities.builtin.provider.message as provider_message
+import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
+import langbot_plugin.api.entities.builtin.resource.tool as resource_tool
+
+
+class GeminiChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """Google Gemini API 请求器"""
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://generativelanguage.googleapis.com/v1beta/openai',
+        'timeout': 120,
+    }
+
+    async def _closure_stream(
+        self,
+        query: pipeline_query.Query,
+        req_messages: list[dict],
+        use_model: requester.RuntimeLLMModel,
+        use_funcs: list[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.MessageChunk:
+        self.client.api_key = use_model.token_mgr.get_token()
+
+        args = {}
+        args['model'] = use_model.model_entity.name
+
+        if use_funcs:
+            tools = await self.ap.tool_mgr.generate_tools_for_openai(use_funcs)
+            if tools:
+                args['tools'] = tools
+
+        # 设置此次请求中的messages
+        messages = req_messages.copy()
+
+        # 检查vision
+        for msg in messages:
+            if 'content' in msg and isinstance(msg['content'], list):
+                for me in msg['content']:
+                    if me['type'] == 'image_base64':
+                        me['image_url'] = {'url': me['image_base64']}
+                        me['type'] = 'image_url'
+                        del me['image_base64']
+
+        args['messages'] = messages
+        args['stream'] = True
+
+        # 流式处理状态
+        # tool_calls_map: dict[str, provider_message.ToolCall] = {}
+        chunk_idx = 0
+        thinking_started = False
+        thinking_ended = False
+        role = 'assistant'  # 默认角色
+        tool_id = ''
+        tool_name = ''
+        # accumulated_reasoning = ''  # 仅用于判断何时结束思维链
+
+        async for chunk in self._req_stream(args, extra_body=extra_args):
+            # 解析 chunk 数据
+
+            if hasattr(chunk, 'choices') and chunk.choices:
+                choice = chunk.choices[0]
+                delta = choice.delta.model_dump() if hasattr(choice, 'delta') else {}
+
+                finish_reason = getattr(choice, 'finish_reason', None)
+            else:
+                delta = {}
+                finish_reason = None
+            # 从第一个 chunk 获取 role，后续使用这个 role
+            if 'role' in delta and delta['role']:
+                role = delta['role']
+
+            # 获取增量内容
+            delta_content = delta.get('content', '')
+            reasoning_content = delta.get('reasoning_content', '')
+
+            # 处理 reasoning_content
+            if reasoning_content:
+                # accumulated_reasoning += reasoning_content
+                # 如果设置了 remove_think，跳过 reasoning_content
+                if remove_think:
+                    chunk_idx += 1
+                    continue
+
+                # 第一次出现 reasoning_content，添加 <think> 开始标签
+                if not thinking_started:
+                    thinking_started = True
+                    delta_content = '<think>\n' + reasoning_content
+                else:
+                    # 继续输出 reasoning_content
+                    delta_content = reasoning_content
+            elif thinking_started and not thinking_ended and delta_content:
+                # reasoning_content 结束，normal content 开始，添加 </think> 结束标签
+                thinking_ended = True
+                delta_content = '\n</think>\n' + delta_content
+
+            # 处理 content 中已有的 <think> 标签（如果需要移除）
+            # if delta_content and remove_think and '<think>' in delta_content:
+            #     import re
+            #
+            #     # 移除 <think> 标签及其内容
+            #     delta_content = re.sub(r'<think>.*?</think>', '', delta_content, flags=re.DOTALL)
+
+            # 处理工具调用增量
+            # delta_tool_calls = None
+            if delta.get('tool_calls'):
+                for tool_call in delta['tool_calls']:
+                    if tool_call['id'] == '' and tool_id == '':
+                        tool_id = str(uuid.uuid4())
+                    if tool_call['function']['name']:
+                        tool_name = tool_call['function']['name']
+                    tool_call['id'] = tool_id
+                    tool_call['function']['name'] = tool_name
+                    if tool_call['type'] is None:
+                        tool_call['type'] = 'function'
+
+            # 跳过空的第一个 chunk（只有 role 没有内容）
+            if chunk_idx == 0 and not delta_content and not reasoning_content and not delta.get('tool_calls'):
+                chunk_idx += 1
+                continue
+            # 构建 MessageChunk - 只包含增量内容
+            chunk_data = {
+                'role': role,
+                'content': delta_content if delta_content else None,
+                'tool_calls': delta.get('tool_calls'),
+                'is_final': bool(finish_reason),
+            }
+
+            # 移除 None 值
+            chunk_data = {k: v for k, v in chunk_data.items() if v is not None}
+
+            yield provider_message.MessageChunk(**chunk_data)
+            chunk_idx += 1
--- a/src/langbot/pkg/provider/modelmgr/requesters/geminichatcmpl.yaml
+++ b/src/langbot/pkg/provider/modelmgr/requesters/geminichatcmpl.yaml
@@ -0,0 +1,31 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: gemini-chat-completions
+  label:
+    en_US: Google Gemini
+    zh_Hans: Google Gemini
+  icon: gemini.svg
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://generativelanguage.googleapis.com/v1beta/openai
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  provider_category: manufacturer
+execution:
+  python:
+    path: ./geminichatcmpl.py
+    attr: GeminiChatCompletions
--- a/src/langbot/pkg/provider/modelmgr/requesters/giteeai.svg
+++ b/src/langbot/pkg/provider/modelmgr/requesters/giteeai.svg
@@ -0,0 +1,3 @@
+<svg width="40" height="40" viewBox="0 0 40 40" fill="none" xmlns="http://www.w3.org/2000/svg">
+<path fill-rule="evenodd" clip-rule="evenodd" d="M25.132 24.3947C25.497 25.7527 25.8984 27.1413 26.3334 28.5834C26.7302 29.8992 25.5459 30.4167 25.0752 29.1758C24.571 27.8466 24.0885 26.523 23.6347 25.1729C21.065 26.4654 18.5025 27.5424 15.5961 28.7541C16.7581 33.0256 17.8309 36.5984 19.4952 39.9935C19.4953 39.9936 19.4953 39.9937 19.4954 39.9938C19.6631 39.9979 19.8313 40 20 40C31.0457 40 40 31.0457 40 20C40 16.0335 38.8453 12.3366 36.8537 9.22729C31.6585 9.69534 27.0513 10.4562 22.8185 11.406C22.8882 12.252 22.9677 13.0739 23.0555 13.855C23.3824 16.7604 23.9112 19.5281 24.6137 22.3836C27.0581 21.2848 29.084 20.3225 30.6816 19.522C32.2154 18.7535 33.6943 18.7062 31.2018 20.6594C29.0388 22.1602 27.0644 23.3566 25.132 24.3947ZM36.1559 8.20846C33.0001 3.89184 28.1561 0.887462 22.5955 0.166882C22.4257 2.86234 22.4785 6.26344 22.681 9.50447C26.7473 8.88859 31.1721 8.46032 36.1559 8.20846ZM19.9369 9.73661e-05C19.7594 2.92694 19.8384 6.65663 20.19 9.91293C17.3748 10.4109 14.7225 11.0064 12.1592 11.7038C12.0486 10.4257 11.9927 9.25764 11.9927 8.24178C11.9927 7.5054 11.3957 6.90844 10.6593 6.90844C9.92296 6.90844 9.32601 7.5054 9.32601 8.24178C9.32601 9.47868 9.42873 10.898 9.61402 12.438C8.33567 12.8278 7.07397 13.2443 5.81918 13.688C5.12493 13.9336 4.76118 14.6954 5.0067 15.3896C5.25223 16.0839 6.01406 16.4476 6.7083 16.2021C7.7931 15.8185 8.88482 15.4388 9.98927 15.0659C10.5222 18.3344 11.3344 21.9428 12.2703 25.4156C12.4336 26.0218 12.6062 26.6262 12.7863 27.2263C9.34168 28.4135 5.82612 29.3782 2.61128 29.8879C0.949407 26.9716 0 23.5967 0 20C0 8.97534 8.92023 0.0341108 19.9369 9.73661e-05ZM4.19152 32.2527C7.45069 36.4516 12.3458 39.3173 17.9204 39.8932C16.5916 37.455 14.9338 33.717 13.5405 29.5901C10.4404 30.7762 7.25883 31.6027 4.19152 32.2527ZM22.9735 23.1135C22.1479 20.41 21.4462 17.5441 20.9225 14.277C20.746 13.5841 20.5918 12.8035 20.4593 11.9636C17.6508 12.6606 14.9992 13.4372 12.4356 14.2598C12.8479 17.4766 13.5448 21.1334 14.5118 24.7218C14.662 25.2792 14.8081 25.8248 14.9514 26.3594L14.9516 26.3603L14.9524 26.3634L14.9526 26.3639L14.973 26.4401C16.1833 25.9872 17.3746 25.5123 18.53 25.0259C20.1235 24.3552 21.6051 23.7165 22.9735 23.1135Z" fill="#141519"/>
+</svg>
--- a/src/langbot/pkg/provider/modelmgr/requesters/giteeaichatcmpl.py
+++ b/src/langbot/pkg/provider/modelmgr/requesters/giteeaichatcmpl.py
@@ -0,0 +1,15 @@
+from __future__ import annotations
+
+
+import typing
+
+from . import ppiochatcmpl
+
+
+class GiteeAIChatCompletions(ppiochatcmpl.PPIOChatCompletions):
+    """Gitee AI ChatCompletions API 请求器"""
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://ai.gitee.com/v1',
+        'timeout': 120,
+    }
--- a/src/langbot/pkg/provider/modelmgr/requesters/giteeaichatcmpl.yaml
+++ b/src/langbot/pkg/provider/modelmgr/requesters/giteeaichatcmpl.yaml
@@ -0,0 +1,32 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: gitee-ai-chat-completions
+  label:
+    en_US: Gitee AI
+    zh_Hans: Gitee AI
+  icon: giteeai.svg
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://ai.gitee.com/v1
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  - text-embedding
+  provider_category: maas
+execution:
+  python:
+    path: ./giteeaichatcmpl.py
+    attr: GiteeAIChatCompletions
--- a/src/langbot/pkg/provider/modelmgr/requesters/jiekouai.png
+++ b/src/langbot/pkg/provider/modelmgr/requesters/jiekouai.png
--- a/src/langbot/pkg/provider/modelmgr/requesters/jiekouaichatcmpl.py
+++ b/src/langbot/pkg/provider/modelmgr/requesters/jiekouaichatcmpl.py
@@ -0,0 +1,208 @@
+from __future__ import annotations
+
+import openai
+import typing
+
+from . import chatcmpl
+from .. import requester
+import openai.types.chat.chat_completion as chat_completion
+import re
+import langbot_plugin.api.entities.builtin.provider.message as provider_message
+import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
+import langbot_plugin.api.entities.builtin.resource.tool as resource_tool
+
+
+class JieKouAIChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """接口 AI ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://api.jiekou.ai/openai',
+        'timeout': 120,
+    }
+
+    is_think: bool = False
+
+    async def _make_msg(
+        self,
+        chat_completion: chat_completion.ChatCompletion,
+        remove_think: bool,
+    ) -> provider_message.Message:
+        chatcmpl_message = chat_completion.choices[0].message.model_dump()
+        # print(chatcmpl_message.keys(), chatcmpl_message.values())
+
+        # 确保 role 字段存在且不为 None
+        if 'role' not in chatcmpl_message or chatcmpl_message['role'] is None:
+            chatcmpl_message['role'] = 'assistant'
+
+        reasoning_content = chatcmpl_message['reasoning_content'] if 'reasoning_content' in chatcmpl_message else None
+
+        # deepseek的reasoner模型
+        chatcmpl_message['content'] = await self._process_thinking_content(
+            chatcmpl_message['content'], reasoning_content, remove_think
+        )
+
+        # 移除 reasoning_content 字段，避免传递给 Message
+        if 'reasoning_content' in chatcmpl_message:
+            del chatcmpl_message['reasoning_content']
+
+        message = provider_message.Message(**chatcmpl_message)
+
+        return message
+
+    async def _process_thinking_content(
+        self,
+        content: str,
+        reasoning_content: str = None,
+        remove_think: bool = False,
+    ) -> tuple[str, str]:
+        """处理思维链内容
+
+        Args:
+            content: 原始内容
+            reasoning_content: reasoning_content 字段内容
+            remove_think: 是否移除思维链
+
+        Returns:
+            处理后的内容
+        """
+        if remove_think:
+            content = re.sub(r'<think>.*?</think>', '', content, flags=re.DOTALL)
+        else:
+            if reasoning_content is not None:
+                content = '<think>\n' + reasoning_content + '\n</think>\n' + content
+        return content
+
+    async def _make_msg_chunk(
+        self,
+        delta: dict[str, typing.Any],
+        idx: int,
+    ) -> provider_message.MessageChunk:
+        # 处理流式chunk和完整响应的差异
+        # print(chat_completion.choices[0])
+
+        # 确保 role 字段存在且不为 None
+        if 'role' not in delta or delta['role'] is None:
+            delta['role'] = 'assistant'
+
+        reasoning_content = delta['reasoning_content'] if 'reasoning_content' in delta else None
+
+        delta['content'] = '' if delta['content'] is None else delta['content']
+        # print(reasoning_content)
+
+        # deepseek的reasoner模型
+
+        if reasoning_content is not None:
+            delta['content'] += reasoning_content
+
+        message = provider_message.MessageChunk(**delta)
+
+        return message
+
+    async def _closure_stream(
+        self,
+        query: pipeline_query.Query,
+        req_messages: list[dict],
+        use_model: requester.RuntimeLLMModel,
+        use_funcs: list[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.Message | typing.AsyncGenerator[provider_message.MessageChunk, None]:
+        self.client.api_key = use_model.token_mgr.get_token()
+
+        args = {}
+        args['model'] = use_model.model_entity.name
+
+        if use_funcs:
+            tools = await self.ap.tool_mgr.generate_tools_for_openai(use_funcs)
+
+            if tools:
+                args['tools'] = tools
+
+        # 设置此次请求中的messages
+        messages = req_messages.copy()
+
+        # 检查vision
+        for msg in messages:
+            if 'content' in msg and isinstance(msg['content'], list):
+                for me in msg['content']:
+                    if me['type'] == 'image_base64':
+                        me['image_url'] = {'url': me['image_base64']}
+                        me['type'] = 'image_url'
+                        del me['image_base64']
+
+        args['messages'] = messages
+        args['stream'] = True
+
+        # tool_calls_map: dict[str, provider_message.ToolCall] = {}
+        chunk_idx = 0
+        thinking_started = False
+        thinking_ended = False
+        role = 'assistant'  # 默认角色
+        async for chunk in self._req_stream(args, extra_body=extra_args):
+            # 解析 chunk 数据
+            if hasattr(chunk, 'choices') and chunk.choices:
+                choice = chunk.choices[0]
+                delta = choice.delta.model_dump() if hasattr(choice, 'delta') else {}
+                finish_reason = getattr(choice, 'finish_reason', None)
+            else:
+                delta = {}
+                finish_reason = None
+
+            # 从第一个 chunk 获取 role，后续使用这个 role
+            if 'role' in delta and delta['role']:
+                role = delta['role']
+
+            # 获取增量内容
+            delta_content = delta.get('content', '')
+            # reasoning_content = delta.get('reasoning_content', '')
+
+            if remove_think:
+                if delta['content'] is not None:
+                    if '<think>' in delta['content'] and not thinking_started and not thinking_ended:
+                        thinking_started = True
+                        continue
+                    elif delta['content'] == r'</think>' and not thinking_ended:
+                        thinking_ended = True
+                        continue
+                    elif thinking_ended and delta['content'] == '\n\n' and thinking_started:
+                        thinking_started = False
+                        continue
+                    elif thinking_started and not thinking_ended:
+                        continue
+
+            # delta_tool_calls = None
+            if delta.get('tool_calls'):
+                for tool_call in delta['tool_calls']:
+                    if tool_call['id'] and tool_call['function']['name']:
+                        tool_id = tool_call['id']
+                        tool_name = tool_call['function']['name']
+
+                    if tool_call['id'] is None:
+                        tool_call['id'] = tool_id
+                    if tool_call['function']['name'] is None:
+                        tool_call['function']['name'] = tool_name
+                    if tool_call['function']['arguments'] is None:
+                        tool_call['function']['arguments'] = ''
+                    if tool_call['type'] is None:
+                        tool_call['type'] = 'function'
+
+            # 跳过空的第一个 chunk（只有 role 没有内容）
+            if chunk_idx == 0 and not delta_content and not delta.get('tool_calls'):
+                chunk_idx += 1
+                continue
+
+            # 构建 MessageChunk - 只包含增量内容
+            chunk_data = {
+                'role': role,
+                'content': delta_content if delta_content else None,
+                'tool_calls': delta.get('tool_calls'),
+                'is_final': bool(finish_reason),
+            }
+
+            # 移除 None 值
+            chunk_data = {k: v for k, v in chunk_data.items() if v is not None}
+
+            yield provider_message.MessageChunk(**chunk_data)
+            chunk_idx += 1
--- a/src/langbot/pkg/provider/modelmgr/requesters/jiekouaichatcmpl.yaml
+++ b/src/langbot/pkg/provider/modelmgr/requesters/jiekouaichatcmpl.yaml
@@ -0,0 +1,39 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: jiekouai-chat-completions
+  label:
+    en_US: JieKou AI
+    zh_Hans: 接口 AI
+  icon: jiekouai.png
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://api.jiekou.ai/openai
+  - name: args
+    label:
+      en_US: Args
+      zh_Hans: 附加参数
+    type: object
+    required: true
+    default: {}
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: int
+    required: true
+    default: 120
+  support_type:
+  - llm
+  - text-embedding
+  provider_category: maas
+execution:
+  python:
+    path: ./jiekouaichatcmpl.py
+    attr: JieKouAIChatCompletions
--- a/src/langbot/pkg/provider/modelmgr/requesters/lmstudio.webp
+++ b/src/langbot/pkg/provider/modelmgr/requesters/lmstudio.webp
--- a/src/langbot/pkg/provider/modelmgr/requesters/lmstudiochatcmpl.py
+++ b/src/langbot/pkg/provider/modelmgr/requesters/lmstudiochatcmpl.py
@@ -0,0 +1,17 @@
+from __future__ import annotations
+
+import typing
+import openai
+
+from . import chatcmpl
+
+
+class LmStudioChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """LMStudio ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'http://127.0.0.1:1234/v1',
+        'timeout': 120,
+    }
--- a/src/langbot/pkg/provider/modelmgr/requesters/lmstudiochatcmpl.yaml
+++ b/src/langbot/pkg/provider/modelmgr/requesters/lmstudiochatcmpl.yaml
@@ -0,0 +1,32 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: lmstudio-chat-completions
+  label:
+    en_US: LM Studio
+    zh_Hans: LM Studio
+  icon: lmstudio.webp
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: http://127.0.0.1:1234/v1
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  - text-embedding
+  provider_category: self-hosted
+execution:
+  python:
+    path: ./lmstudiochatcmpl.py
+    attr: LmStudioChatCompletions
--- a/src/langbot/pkg/provider/modelmgr/requesters/modelscope.svg
+++ b/src/langbot/pkg/provider/modelmgr/requesters/modelscope.svg
@@ -0,0 +1 @@
+<svg id="_层_2" xmlns="http://www.w3.org/2000/svg" viewBox="0 0 274.37 172.76"><defs><style>.cls-2{fill:#36cfd1}.cls-3{fill:#624aff}</style></defs><g id="_层_1-2"><path class="cls-3" d="M24.78 73.55h25.65V99.2H24.78zm99.14 25.66h25.65v25.65h-25.65zm76.95 25.65h-25.65v22.19h47.84V99.21h-22.19v25.65z"/><path class="cls-2" d="M149.57 73.55h25.65V99.2h-25.65zM24.78 47.9h25.65v25.65H24.78z"/><path class="cls-3" d="M223.06 73.55h25.65V99.2h-25.65z"/><path class="cls-2" d="M223.06 47.9h25.65v25.65h-25.65z"/><path class="cls-3" d="M175.22 25.71V47.9h25.65v25.65h22.19V25.71h-47.84z"/><path class="cls-2" d="M98.27 73.55h25.65V99.2H98.27z"/><path class="cls-3" d="M72.62 47.9h25.65V25.71H50.43v47.84h22.19V47.9zm0 51.31H50.43v47.84h47.84v-22.19H72.62V99.21z"/><path style="fill:none" d="M0 0h274.37v172.76H0z"/></g></svg>
--- a/src/langbot/pkg/provider/modelmgr/requesters/modelscopechatcmpl.py
+++ b/src/langbot/pkg/provider/modelmgr/requesters/modelscopechatcmpl.py
@@ -0,0 +1,389 @@
+from __future__ import annotations
+
+import asyncio
+import typing
+
+import openai
+import openai.types.chat.chat_completion as chat_completion
+import httpx
+
+from .. import entities, errors, requester
+import langbot_plugin.api.entities.builtin.resource.tool as resource_tool
+import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
+import langbot_plugin.api.entities.builtin.provider.message as provider_message
+
+
+class ModelScopeChatCompletions(requester.ProviderAPIRequester):
+    """ModelScope ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://api-inference.modelscope.cn/v1',
+        'timeout': 120,
+    }
+
+    async def initialize(self):
+        self.client = openai.AsyncClient(
+            api_key='',
+            base_url=self.requester_cfg['base_url'],
+            timeout=self.requester_cfg['timeout'],
+            http_client=httpx.AsyncClient(trust_env=True, timeout=self.requester_cfg['timeout']),
+        )
+
+    async def _req(
+        self,
+        query: pipeline_query.Query,
+        args: dict,
+        extra_body: dict = {},
+        remove_think: bool = False,
+    ) -> list[dict[str, typing.Any]]:
+        args['stream'] = True
+
+        chunk = None
+
+        pending_content = ''
+
+        tool_calls = []
+
+        resp_gen: openai.AsyncStream = await self.client.chat.completions.create(**args, extra_body=extra_body)
+
+        chunk_idx = 0
+        thinking_started = False
+        thinking_ended = False
+        tool_id = ''
+        tool_name = ''
+        message_delta = {}
+        async for chunk in resp_gen:
+            if not chunk or not chunk.id or not chunk.choices or not chunk.choices[0] or not chunk.choices[0].delta:
+                continue
+
+            delta = chunk.choices[0].delta.model_dump() if hasattr(chunk.choices[0], 'delta') else {}
+            reasoning_content = delta.get('reasoning_content')
+            # 处理 reasoning_content
+            if reasoning_content:
+                # accumulated_reasoning += reasoning_content
+                # 如果设置了 remove_think，跳过 reasoning_content
+                if remove_think:
+                    chunk_idx += 1
+                    continue
+
+                # 第一次出现 reasoning_content，添加 <think> 开始标签
+                if not thinking_started:
+                    thinking_started = True
+                    pending_content += '<think>\n' + reasoning_content
+                else:
+                    # 继续输出 reasoning_content
+                    pending_content += reasoning_content
+            elif thinking_started and not thinking_ended and delta.get('content'):
+                # reasoning_content 结束，normal content 开始，添加 </think> 结束标签
+                thinking_ended = True
+                pending_content += '\n</think>\n' + delta.get('content')
+
+            if delta.get('content') is not None:
+                pending_content += delta.get('content')
+
+            if delta.get('tool_calls') is not None:
+                for tool_call in delta.get('tool_calls'):
+                    if tool_call['id'] != '':
+                        tool_id = tool_call['id']
+                    if tool_call['function']['name'] is not None:
+                        tool_name = tool_call['function']['name']
+                    if tool_call['function']['arguments'] is None:
+                        continue
+                    tool_call['id'] = tool_id
+                    tool_call['name'] = tool_name
+                    for tc in tool_calls:
+                        if tc['index'] == tool_call['index']:
+                            tc['function']['arguments'] += tool_call['function']['arguments']
+                            break
+                    else:
+                        tool_calls.append(tool_call)
+
+            if chunk.choices[0].finish_reason is not None:
+                break
+        message_delta['content'] = pending_content
+        message_delta['role'] = 'assistant'
+
+        message_delta['tool_calls'] = tool_calls if tool_calls else None
+        return [message_delta]
+
+    async def _make_msg(
+        self,
+        chat_completion: list[dict[str, typing.Any]],
+    ) -> provider_message.Message:
+        chatcmpl_message = chat_completion[0]
+
+        # 确保 role 字段存在且不为 None
+        if 'role' not in chatcmpl_message or chatcmpl_message['role'] is None:
+            chatcmpl_message['role'] = 'assistant'
+
+        message = provider_message.Message(**chatcmpl_message)
+
+        return message
+
+    async def _closure(
+        self,
+        query: pipeline_query.Query,
+        req_messages: list[dict],
+        use_model: requester.RuntimeLLMModel,
+        use_funcs: list[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.Message:
+        self.client.api_key = use_model.token_mgr.get_token()
+
+        args = {}
+        args['model'] = use_model.model_entity.name
+
+        if use_funcs:
+            tools = await self.ap.tool_mgr.generate_tools_for_openai(use_funcs)
+
+            if tools:
+                args['tools'] = tools
+
+        # 设置此次请求中的messages
+        messages = req_messages.copy()
+
+        # 检查vision
+        for msg in messages:
+            if 'content' in msg and isinstance(msg['content'], list):
+                for me in msg['content']:
+                    if me['type'] == 'image_base64':
+                        me['image_url'] = {'url': me['image_base64']}
+                        me['type'] = 'image_url'
+                        del me['image_base64']
+
+        args['messages'] = messages
+
+        # 发送请求
+        resp = await self._req(query, args, extra_body=extra_args, remove_think=remove_think)
+
+        # 处理请求结果
+        message = await self._make_msg(resp)
+
+        return message
+
+    async def _req_stream(
+        self,
+        args: dict,
+        extra_body: dict = {},
+    ) -> chat_completion.ChatCompletion:
+        async for chunk in await self.client.chat.completions.create(**args, extra_body=extra_body):
+            yield chunk
+
+    async def _closure_stream(
+        self,
+        query: pipeline_query.Query,
+        req_messages: list[dict],
+        use_model: requester.RuntimeLLMModel,
+        use_funcs: list[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.Message | typing.AsyncGenerator[provider_message.MessageChunk, None]:
+        self.client.api_key = use_model.token_mgr.get_token()
+
+        args = {}
+        args['model'] = use_model.model_entity.name
+
+        if use_funcs:
+            tools = await self.ap.tool_mgr.generate_tools_for_openai(use_funcs)
+
+            if tools:
+                args['tools'] = tools
+
+        # 设置此次请求中的messages
+        messages = req_messages.copy()
+
+        # 检查vision
+        for msg in messages:
+            if 'content' in msg and isinstance(msg['content'], list):
+                for me in msg['content']:
+                    if me['type'] == 'image_base64':
+                        me['image_url'] = {'url': me['image_base64']}
+                        me['type'] = 'image_url'
+                        del me['image_base64']
+
+        args['messages'] = messages
+        args['stream'] = True
+
+        # 流式处理状态
+        # tool_calls_map: dict[str, provider_message.ToolCall] = {}
+        chunk_idx = 0
+        thinking_started = False
+        thinking_ended = False
+        role = 'assistant'  # 默认角色
+        # accumulated_reasoning = ''  # 仅用于判断何时结束思维链
+
+        async for chunk in self._req_stream(args, extra_body=extra_args):
+            # 解析 chunk 数据
+            if hasattr(chunk, 'choices') and chunk.choices:
+                choice = chunk.choices[0]
+                delta = choice.delta.model_dump() if hasattr(choice, 'delta') else {}
+                finish_reason = getattr(choice, 'finish_reason', None)
+            else:
+                delta = {}
+                finish_reason = None
+
+            # 从第一个 chunk 获取 role，后续使用这个 role
+            if 'role' in delta and delta['role']:
+                role = delta['role']
+
+            # 获取增量内容
+            delta_content = delta.get('content', '')
+            reasoning_content = delta.get('reasoning_content', '')
+
+            # 处理 reasoning_content
+            if reasoning_content:
+                # accumulated_reasoning += reasoning_content
+                # 如果设置了 remove_think，跳过 reasoning_content
+                if remove_think:
+                    chunk_idx += 1
+                    continue
+
+                # 第一次出现 reasoning_content，添加 <think> 开始标签
+                if not thinking_started:
+                    thinking_started = True
+                    delta_content = '<think>\n' + reasoning_content
+                else:
+                    # 继续输出 reasoning_content
+                    delta_content = reasoning_content
+            elif thinking_started and not thinking_ended and delta_content:
+                # reasoning_content 结束，normal content 开始，添加 </think> 结束标签
+                thinking_ended = True
+                delta_content = '\n</think>\n' + delta_content
+
+            # 处理 content 中已有的 <think> 标签（如果需要移除）
+            # if delta_content and remove_think and '<think>' in delta_content:
+            #     import re
+            #
+            #     # 移除 <think> 标签及其内容
+            #     delta_content = re.sub(r'<think>.*?</think>', '', delta_content, flags=re.DOTALL)
+
+            # 处理工具调用增量
+            if delta.get('tool_calls'):
+                for tool_call in delta['tool_calls']:
+                    if tool_call['id'] != '':
+                        tool_id = tool_call['id']
+                    if tool_call['function']['name'] is not None:
+                        tool_name = tool_call['function']['name']
+
+                    if tool_call['type'] is None:
+                        tool_call['type'] = 'function'
+                    tool_call['id'] = tool_id
+                    tool_call['function']['name'] = tool_name
+                    tool_call['function']['arguments'] = (
+                        '' if tool_call['function']['arguments'] is None else tool_call['function']['arguments']
+                    )
+
+            # 跳过空的第一个 chunk（只有 role 没有内容）
+            if chunk_idx == 0 and not delta_content and not reasoning_content and not delta.get('tool_calls'):
+                chunk_idx += 1
+                continue
+
+            # 构建 MessageChunk - 只包含增量内容
+            chunk_data = {
+                'role': role,
+                'content': delta_content if delta_content else None,
+                'tool_calls': delta.get('tool_calls'),
+                'is_final': bool(finish_reason),
+            }
+
+            # 移除 None 值
+            chunk_data = {k: v for k, v in chunk_data.items() if v is not None}
+
+            yield provider_message.MessageChunk(**chunk_data)
+            chunk_idx += 1
+            # return
+
+    async def invoke_llm(
+        self,
+        query: pipeline_query.Query,
+        model: entities.LLMModelInfo,
+        messages: typing.List[provider_message.Message],
+        funcs: typing.List[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.Message:
+        req_messages = []  # req_messages 仅用于类内，外部同步由 query.messages 进行
+        for m in messages:
+            msg_dict = m.dict(exclude_none=True)
+            content = msg_dict.get('content')
+            if isinstance(content, list):
+                # 检查 content 列表中是否每个部分都是文本
+                if all(isinstance(part, dict) and part.get('type') == 'text' for part in content):
+                    # 将所有文本部分合并为一个字符串
+                    msg_dict['content'] = '\n'.join(part['text'] for part in content)
+            req_messages.append(msg_dict)
+
+        try:
+            return await self._closure(
+                query=query,
+                req_messages=req_messages,
+                use_model=model,
+                use_funcs=funcs,
+                extra_args=extra_args,
+                remove_think=remove_think,
+            )
+        except asyncio.TimeoutError:
+            raise errors.RequesterError('请求超时')
+        except openai.BadRequestError as e:
+            if 'context_length_exceeded' in e.message:
+                raise errors.RequesterError(f'上文过长，请重置会话: {e.message}')
+            else:
+                raise errors.RequesterError(f'请求参数错误: {e.message}')
+        except openai.AuthenticationError as e:
+            raise errors.RequesterError(f'无效的 api-key: {e.message}')
+        except openai.NotFoundError as e:
+            raise errors.RequesterError(f'请求路径错误: {e.message}')
+        except openai.RateLimitError as e:
+            raise errors.RequesterError(f'请求过于频繁或余额不足: {e.message}')
+        except openai.APIError as e:
+            raise errors.RequesterError(f'请求错误: {e.message}')
+
+    async def invoke_llm_stream(
+        self,
+        query: pipeline_query.Query,
+        model: requester.RuntimeLLMModel,
+        messages: typing.List[provider_message.Message],
+        funcs: typing.List[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.MessageChunk:
+        req_messages = []  # req_messages 仅用于类内，外部同步由 query.messages 进行
+        for m in messages:
+            msg_dict = m.dict(exclude_none=True)
+            content = msg_dict.get('content')
+            if isinstance(content, list):
+                # 检查 content 列表中是否每个部分都是文本
+                if all(isinstance(part, dict) and part.get('type') == 'text' for part in content):
+                    # 将所有文本部分合并为一个字符串
+                    msg_dict['content'] = '\n'.join(part['text'] for part in content)
+            req_messages.append(msg_dict)
+
+        try:
+            async for item in self._closure_stream(
+                query=query,
+                req_messages=req_messages,
+                use_model=model,
+                use_funcs=funcs,
+                extra_args=extra_args,
+                remove_think=remove_think,
+            ):
+                yield item
+
+        except asyncio.TimeoutError:
+            raise errors.RequesterError('请求超时')
+        except openai.BadRequestError as e:
+            if 'context_length_exceeded' in e.message:
+                raise errors.RequesterError(f'上文过长，请重置会话: {e.message}')
+            else:
+                raise errors.RequesterError(f'请求参数错误: {e.message}')
+        except openai.AuthenticationError as e:
+            raise errors.RequesterError(f'无效的 api-key: {e.message}')
+        except openai.NotFoundError as e:
+            raise errors.RequesterError(f'请求路径错误: {e.message}')
+        except openai.RateLimitError as e:
+            raise errors.RequesterError(f'请求过于频繁或余额不足: {e.message}')
+        except openai.APIError as e:
+            raise errors.RequesterError(f'请求错误: {e.message}')
--- a/src/langbot/pkg/provider/modelmgr/requesters/modelscopechatcmpl.yaml
+++ b/src/langbot/pkg/provider/modelmgr/requesters/modelscopechatcmpl.yaml
@@ -0,0 +1,38 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: modelscope-chat-completions
+  label:
+    en_US: ModelScope
+    zh_Hans: 魔搭社区
+  icon: modelscope.svg
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://api-inference.modelscope.cn/v1
+  - name: args
+    label:
+      en_US: Args
+      zh_Hans: 附加参数
+    type: object
+    required: true
+    default: {}
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: int
+    required: true
+    default: 120
+  support_type:
+  - llm
+  provider_category: maas
+execution:
+  python:
+    path: ./modelscopechatcmpl.py
+    attr: ModelScopeChatCompletions
--- a/src/langbot/pkg/provider/modelmgr/requesters/moonshot.png
+++ b/src/langbot/pkg/provider/modelmgr/requesters/moonshot.png
--- a/src/langbot/pkg/provider/modelmgr/requesters/moonshotchatcmpl.py
+++ b/src/langbot/pkg/provider/modelmgr/requesters/moonshotchatcmpl.py
@@ -0,0 +1,60 @@
+from __future__ import annotations
+
+import typing
+
+
+from . import chatcmpl
+from .. import requester
+import langbot_plugin.api.entities.builtin.resource.tool as resource_tool
+import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
+import langbot_plugin.api.entities.builtin.provider.message as provider_message
+
+
+class MoonshotChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """Moonshot ChatCompletion API 请求器"""
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://api.moonshot.cn/v1',
+        'timeout': 120,
+    }
+
+    async def _closure(
+        self,
+        query: pipeline_query.Query,
+        req_messages: list[dict],
+        use_model: requester.RuntimeLLMModel,
+        use_funcs: list[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.Message:
+        self.client.api_key = use_model.token_mgr.get_token()
+
+        args = {}
+        args['model'] = use_model.model_entity.name
+
+        if use_funcs:
+            tools = await self.ap.tool_mgr.generate_tools_for_openai(use_funcs)
+
+            if tools:
+                args['tools'] = tools
+
+        # 设置此次请求中的messages
+        messages = req_messages
+
+        # deepseek 不支持多模态，把content都转换成纯文字
+        for m in messages:
+            if 'content' in m and isinstance(m['content'], list):
+                m['content'] = ' '.join([c['text'] for c in m['content']])
+
+        # 删除空的，不知道干嘛的，直接删了。
+        # messages = [m for m in messages if m["content"].strip() != "" and ('tool_calls' not in m or not m['tool_calls'])]
+
+        args['messages'] = messages
+
+        # 发送请求
+        resp = await self._req(args, extra_body=extra_args)
+
+        # 处理请求结果
+        message = await self._make_msg(resp, remove_think)
+
+        return message
--- a/src/langbot/pkg/provider/modelmgr/requesters/moonshotchatcmpl.yaml
+++ b/src/langbot/pkg/provider/modelmgr/requesters/moonshotchatcmpl.yaml
@@ -0,0 +1,31 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: moonshot-chat-completions
+  label:
+    en_US: Moonshot
+    zh_Hans: 月之暗面
+  icon: moonshot.png
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://api.moonshot.ai/v1
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  provider_category: manufacturer
+execution:
+  python:
+    path: ./moonshotchatcmpl.py
+    attr: MoonshotChatCompletions
--- a/src/langbot/pkg/provider/modelmgr/requesters/newapi.png
+++ b/src/langbot/pkg/provider/modelmgr/requesters/newapi.png
--- a/src/langbot/pkg/provider/modelmgr/requesters/newapichatcmpl.py
+++ b/src/langbot/pkg/provider/modelmgr/requesters/newapichatcmpl.py
@@ -0,0 +1,17 @@
+from __future__ import annotations
+
+import typing
+import openai
+
+from . import chatcmpl
+
+
+class NewAPIChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """New API ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'http://localhost:3000/v1',
+        'timeout': 120,
+    }
--- a/src/langbot/pkg/provider/modelmgr/requesters/newapichatcmpl.yaml
+++ b/src/langbot/pkg/provider/modelmgr/requesters/newapichatcmpl.yaml
@@ -0,0 +1,32 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: new-api-chat-completions
+  label:
+    en_US: New API
+    zh_Hans: New API
+  icon: newapi.png
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: http://localhost:3000/v1
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  - text-embedding
+  provider_category: maas
+execution:
+  python:
+    path: ./newapichatcmpl.py
+    attr: NewAPIChatCompletions
--- a/src/langbot/pkg/provider/modelmgr/requesters/ollama.svg
+++ b/src/langbot/pkg/provider/modelmgr/requesters/ollama.svg
--- a/src/langbot/pkg/provider/modelmgr/requesters/ollamachat.py
+++ b/src/langbot/pkg/provider/modelmgr/requesters/ollamachat.py
@@ -0,0 +1,148 @@
+from __future__ import annotations
+
+import asyncio
+import os
+import typing
+from typing import Union, Mapping, Any, AsyncIterator
+import uuid
+import json
+
+import ollama
+
+from .. import errors, requester
+import langbot_plugin.api.entities.builtin.resource.tool as resource_tool
+import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
+import langbot_plugin.api.entities.builtin.provider.message as provider_message
+
+REQUESTER_NAME: str = 'ollama-chat'
+
+
+class OllamaChatCompletions(requester.ProviderAPIRequester):
+    """Ollama平台 ChatCompletion API请求器"""
+
+    client: ollama.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'http://127.0.0.1:11434',
+        'timeout': 120,
+    }
+
+    async def initialize(self):
+        os.environ['OLLAMA_HOST'] = self.requester_cfg['base_url']
+        self.client = ollama.AsyncClient(timeout=self.requester_cfg['timeout'])
+
+    async def _req(
+        self,
+        args: dict,
+    ) -> Union[Mapping[str, Any], AsyncIterator[Mapping[str, Any]]]:
+        return await self.client.chat(**args)
+
+    async def _closure(
+        self,
+        query: pipeline_query.Query,
+        req_messages: list[dict],
+        use_model: requester.RuntimeLLMModel,
+        use_funcs: list[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.Message:
+        args = extra_args.copy()
+        args['model'] = use_model.model_entity.name
+
+        messages: list[dict] = req_messages.copy()
+        for msg in messages:
+            if 'content' in msg and isinstance(msg['content'], list):
+                text_content: list = []
+                image_urls: list = []
+                for me in msg['content']:
+                    if me['type'] == 'text':
+                        text_content.append(me['text'])
+                    elif me['type'] == 'image_base64':
+                        image_urls.append(me['image_base64'])
+
+                msg['content'] = '\n'.join(text_content)
+                msg['images'] = [url.split(',')[1] for url in image_urls]
+            if 'tool_calls' in msg:  # LangBot 内部以 str 存储 tool_calls 的参数，这里需要转换为 dict
+                for tool_call in msg['tool_calls']:
+                    tool_call['function']['arguments'] = json.loads(tool_call['function']['arguments'])
+        args['messages'] = messages
+
+        args['tools'] = []
+        if use_funcs:
+            tools = await self.ap.tool_mgr.generate_tools_for_openai(use_funcs)
+            if tools:
+                args['tools'] = tools
+
+        resp = await self._req(args)
+        message: provider_message.Message = await self._make_msg(resp)
+        return message
+
+    async def _make_msg(self, chat_completions: ollama.ChatResponse) -> provider_message.Message:
+        message: ollama.Message = chat_completions.message
+        if message is None:
+            raise ValueError("chat_completions must contain a 'message' field")
+
+        ret_msg: provider_message.Message = None
+
+        if message.content is not None:
+            ret_msg = provider_message.Message(role='assistant', content=message.content)
+        if message.tool_calls is not None and len(message.tool_calls) > 0:
+            tool_calls: list[provider_message.ToolCall] = []
+
+            for tool_call in message.tool_calls:
+                tool_calls.append(
+                    provider_message.ToolCall(
+                        id=uuid.uuid4().hex,
+                        type='function',
+                        function=provider_message.FunctionCall(
+                            name=tool_call.function.name,
+                            arguments=json.dumps(tool_call.function.arguments),
+                        ),
+                    )
+                )
+            ret_msg.tool_calls = tool_calls
+
+        return ret_msg
+
+    async def invoke_llm(
+        self,
+        query: pipeline_query.Query,
+        model: requester.RuntimeLLMModel,
+        messages: typing.List[provider_message.Message],
+        funcs: typing.List[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.Message:
+        req_messages: list = []
+        for m in messages:
+            msg_dict: dict = m.dict(exclude_none=True)
+            content: Any = msg_dict.get('content')
+            if isinstance(content, list):
+                if all(isinstance(part, dict) and part.get('type') == 'text' for part in content):
+                    msg_dict['content'] = '\n'.join(part['text'] for part in content)
+            req_messages.append(msg_dict)
+        try:
+            return await self._closure(
+                query=query,
+                req_messages=req_messages,
+                use_model=model,
+                use_funcs=funcs,
+                extra_args=extra_args,
+                remove_think=remove_think,
+            )
+        except asyncio.TimeoutError:
+            raise errors.RequesterError('请求超时')
+
+    async def invoke_embedding(
+        self,
+        model: requester.RuntimeEmbeddingModel,
+        input_text: list[str],
+        extra_args: dict[str, typing.Any] = {},
+    ) -> list[list[float]]:
+        return (
+            await self.client.embed(
+                model=model.model_entity.name,
+                input=input_text,
+                **extra_args,
+            )
+        ).embeddings
--- a/src/langbot/pkg/provider/modelmgr/requesters/ollamachat.yaml
+++ b/src/langbot/pkg/provider/modelmgr/requesters/ollamachat.yaml
@@ -0,0 +1,32 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: ollama-chat
+  label:
+    en_US: Ollama
+    zh_Hans: Ollama
+  icon: ollama.svg
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: http://127.0.0.1:11434
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  - text-embedding
+  provider_category: self-hosted
+execution:
+  python:
+    path: ./ollamachat.py
+    attr: OllamaChatCompletions
--- a/src/langbot/pkg/provider/modelmgr/requesters/openai.svg
+++ b/src/langbot/pkg/provider/modelmgr/requesters/openai.svg
--- a/src/langbot/pkg/provider/modelmgr/requesters/openrouter.svg
+++ b/src/langbot/pkg/provider/modelmgr/requesters/openrouter.svg
@@ -0,0 +1,10 @@
+<svg width="25" height="21" viewBox="0 0 25 21" fill="none" xmlns="http://www.w3.org/2000/svg">
+<path d="M1.05858 10.1738C1.76158 10.1738 4.47988 9.56715 5.88589 8.77041C7.2919 7.97367 7.2919 7.97367 10.1977 5.91152C13.8766 3.30069 16.4779 4.17486 20.7428 4.17486" fill="black"/>
+<path fill-rule="evenodd" clip-rule="evenodd" d="M11.4182 7.63145L11.3787 7.65951C8.50565 9.69845 8.42504 9.75566 6.92566 10.6053C5.98567 11.138 4.74704 11.5436 3.75151 11.8089C2.80313 12.0615 1.71203 12.2829 1.05858 12.2829V8.06483C1.05075 8.06483 1.05422 8.06445 1.06984 8.06276C1.11491 8.05788 1.26116 8.04203 1.52896 7.9926C1.84599 7.9341 2.24205 7.84582 2.6657 7.73296C3.55657 7.49564 4.3801 7.1996 4.84612 6.93552C4.88175 6.91533 4.91635 6.89573 4.95001 6.87666C6.15007 6.19693 6.15657 6.19325 8.97708 4.1916C12.5199 1.67735 15.5815 1.83587 18.5849 1.99138C19.3056 2.0287 20.0229 2.06584 20.7428 2.06584V6.28388C19.6102 6.28388 18.6583 6.24193 17.8263 6.20527C15.1245 6.08621 13.685 6.02278 11.4182 7.63145Z" fill="black"/>
+<path d="M24.8671 4.20087L17.6613 8.36117V0.0405881L24.8671 4.20087Z" fill="black"/>
+<path fill-rule="evenodd" clip-rule="evenodd" d="M17.6378 0L24.9139 4.20087L17.6378 8.40176V0ZM17.6847 0.0811762V8.32058L24.8202 4.20087L17.6847 0.0811762Z" fill="black"/>
+<path d="M0.917975 10.1764C1.62098 10.1764 4.33927 10.7831 5.74529 11.5799C7.1513 12.3766 7.1513 12.3766 10.0571 14.4388C13.736 17.0496 16.3373 16.1754 20.6022 16.1754" fill="black"/>
+<path fill-rule="evenodd" clip-rule="evenodd" d="M0.929234 12.2875C0.913615 12.2858 0.910145 12.2854 0.917975 12.2854V8.06741C1.57142 8.06741 2.66253 8.28878 3.61091 8.54142C4.60644 8.80663 5.84507 9.21231 6.78506 9.74497C8.28444 10.5946 8.36505 10.6518 11.2381 12.6908L11.2776 12.7188C13.5444 14.3275 14.9839 14.2641 17.6857 14.145C18.5177 14.1083 19.4696 14.0664 20.6022 14.0664V18.2844C19.8823 18.2844 19.165 18.3216 18.4443 18.3589C15.4409 18.5144 12.3793 18.6729 8.83648 16.1587C6.01597 14.157 6.00947 14.1533 4.80941 13.4736C4.77575 13.4545 4.74115 13.4349 4.70551 13.4148C4.2395 13.1507 3.41597 12.8546 2.5251 12.6173C2.10145 12.5045 1.70538 12.4162 1.38836 12.3577C1.12056 12.3083 0.974309 12.2924 0.929234 12.2875Z" fill="black"/>
+<path d="M24.7265 16.1494L17.5207 11.9892V20.3097L24.7265 16.1494Z" fill="black"/>
+<path fill-rule="evenodd" clip-rule="evenodd" d="M17.4972 11.9486L24.7733 16.1494L17.4972 20.3503V11.9486ZM17.5441 12.0297V20.2691L24.6796 16.1494L17.5441 12.0297Z" fill="black"/>
+</svg>
--- a/src/langbot/pkg/provider/modelmgr/requesters/openrouterchatcmpl.py
+++ b/src/langbot/pkg/provider/modelmgr/requesters/openrouterchatcmpl.py
@@ -0,0 +1,17 @@
+from __future__ import annotations
+
+import typing
+import openai
+
+from . import modelscopechatcmpl
+
+
+class OpenRouterChatCompletions(modelscopechatcmpl.ModelScopeChatCompletions):
+    """OpenRouter ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://openrouter.ai/api/v1',
+        'timeout': 120,
+    }
--- a/src/langbot/pkg/provider/modelmgr/requesters/openrouterchatcmpl.yaml
+++ b/src/langbot/pkg/provider/modelmgr/requesters/openrouterchatcmpl.yaml
@@ -0,0 +1,32 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: openrouter-chat-completions
+  label:
+    en_US: OpenRouter
+    zh_Hans: OpenRouter
+  icon: openrouter.svg
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://openrouter.ai/api/v1
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  - text-embedding
+  provider_category: maas
+execution:
+  python:
+    path: ./openrouterchatcmpl.py
+    attr: OpenRouterChatCompletions
--- a/src/langbot/pkg/provider/modelmgr/requesters/ppio.svg
+++ b/src/langbot/pkg/provider/modelmgr/requesters/ppio.svg
@@ -0,0 +1,3 @@
+<svg width="60" height="60" viewBox="0 0 60 60" fill="none" xmlns="http://www.w3.org/2000/svg">
+<path d="M29.7888 0.215881C13.3449 0.215881 0 13.5422 0 29.986C0 38.0916 3.24782 45.4527 8.51506 50.8223V30.0139C8.51506 24.3372 10.7299 18.9769 14.7408 14.966C18.7704 10.9365 24.112 8.74025 29.7981 8.74025H29.9749L29.7888 8.75886C41.5423 8.75886 51.0718 18.2883 51.0718 30.0326C51.0718 31.0562 50.9973 32.0613 50.8577 33.057L38.8343 20.9964C36.4333 18.5954 33.2134 17.2646 29.8074 17.2646C26.4013 17.2646 23.1907 18.5954 20.7805 20.9964C18.3609 23.4159 17.0394 26.6172 17.0394 30.0326C17.0394 33.4479 18.3702 36.6492 20.7805 39.0688C23.1814 41.4697 26.4013 42.8005 29.8074 42.8005C33.2134 42.8005 36.424 41.4697 38.8343 39.0688C41.077 36.826 42.3706 33.8946 42.5474 30.7584L49.6014 37.8403C46.4839 45.7319 38.797 51.3249 29.7981 51.3249C25.1357 51.3249 20.6874 49.8359 17.0301 47.072V56.9178C20.9014 58.7604 25.2195 59.7841 29.7794 59.7841C46.2233 59.7841 59.5682 46.4578 59.5682 30.0139C59.5868 13.5515 46.2512 0.225187 29.7981 0.225187L29.7888 0.215881Z" fill="#0062E2"/>
+</svg>
--- a/src/langbot/pkg/provider/modelmgr/requesters/ppiochatcmpl.py
+++ b/src/langbot/pkg/provider/modelmgr/requesters/ppiochatcmpl.py
@@ -0,0 +1,208 @@
+from __future__ import annotations
+
+import openai
+import typing
+
+from . import chatcmpl
+from .. import requester
+import openai.types.chat.chat_completion as chat_completion
+import re
+import langbot_plugin.api.entities.builtin.provider.message as provider_message
+import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
+import langbot_plugin.api.entities.builtin.resource.tool as resource_tool
+
+
+class PPIOChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """欧派云 ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://api.ppinfra.com/v3/openai',
+        'timeout': 120,
+    }
+
+    is_think: bool = False
+
+    async def _make_msg(
+        self,
+        chat_completion: chat_completion.ChatCompletion,
+        remove_think: bool,
+    ) -> provider_message.Message:
+        chatcmpl_message = chat_completion.choices[0].message.model_dump()
+        # print(chatcmpl_message.keys(), chatcmpl_message.values())
+
+        # 确保 role 字段存在且不为 None
+        if 'role' not in chatcmpl_message or chatcmpl_message['role'] is None:
+            chatcmpl_message['role'] = 'assistant'
+
+        reasoning_content = chatcmpl_message['reasoning_content'] if 'reasoning_content' in chatcmpl_message else None
+
+        # deepseek的reasoner模型
+        chatcmpl_message['content'] = await self._process_thinking_content(
+            chatcmpl_message['content'], reasoning_content, remove_think
+        )
+
+        # 移除 reasoning_content 字段，避免传递给 Message
+        if 'reasoning_content' in chatcmpl_message:
+            del chatcmpl_message['reasoning_content']
+
+        message = provider_message.Message(**chatcmpl_message)
+
+        return message
+
+    async def _process_thinking_content(
+        self,
+        content: str,
+        reasoning_content: str = None,
+        remove_think: bool = False,
+    ) -> tuple[str, str]:
+        """处理思维链内容
+
+        Args:
+            content: 原始内容
+            reasoning_content: reasoning_content 字段内容
+            remove_think: 是否移除思维链
+
+        Returns:
+            处理后的内容
+        """
+        if remove_think:
+            content = re.sub(r'<think>.*?</think>', '', content, flags=re.DOTALL)
+        else:
+            if reasoning_content is not None:
+                content = '<think>\n' + reasoning_content + '\n</think>\n' + content
+        return content
+
+    async def _make_msg_chunk(
+        self,
+        delta: dict[str, typing.Any],
+        idx: int,
+    ) -> provider_message.MessageChunk:
+        # 处理流式chunk和完整响应的差异
+        # print(chat_completion.choices[0])
+
+        # 确保 role 字段存在且不为 None
+        if 'role' not in delta or delta['role'] is None:
+            delta['role'] = 'assistant'
+
+        reasoning_content = delta['reasoning_content'] if 'reasoning_content' in delta else None
+
+        delta['content'] = '' if delta['content'] is None else delta['content']
+        # print(reasoning_content)
+
+        # deepseek的reasoner模型
+
+        if reasoning_content is not None:
+            delta['content'] += reasoning_content
+
+        message = provider_message.MessageChunk(**delta)
+
+        return message
+
+    async def _closure_stream(
+        self,
+        query: pipeline_query.Query,
+        req_messages: list[dict],
+        use_model: requester.RuntimeLLMModel,
+        use_funcs: list[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.Message | typing.AsyncGenerator[provider_message.MessageChunk, None]:
+        self.client.api_key = use_model.token_mgr.get_token()
+
+        args = {}
+        args['model'] = use_model.model_entity.name
+
+        if use_funcs:
+            tools = await self.ap.tool_mgr.generate_tools_for_openai(use_funcs)
+
+            if tools:
+                args['tools'] = tools
+
+        # 设置此次请求中的messages
+        messages = req_messages.copy()
+
+        # 检查vision
+        for msg in messages:
+            if 'content' in msg and isinstance(msg['content'], list):
+                for me in msg['content']:
+                    if me['type'] == 'image_base64':
+                        me['image_url'] = {'url': me['image_base64']}
+                        me['type'] = 'image_url'
+                        del me['image_base64']
+
+        args['messages'] = messages
+        args['stream'] = True
+
+        # tool_calls_map: dict[str, provider_message.ToolCall] = {}
+        chunk_idx = 0
+        thinking_started = False
+        thinking_ended = False
+        role = 'assistant'  # 默认角色
+        async for chunk in self._req_stream(args, extra_body=extra_args):
+            # 解析 chunk 数据
+            if hasattr(chunk, 'choices') and chunk.choices:
+                choice = chunk.choices[0]
+                delta = choice.delta.model_dump() if hasattr(choice, 'delta') else {}
+                finish_reason = getattr(choice, 'finish_reason', None)
+            else:
+                delta = {}
+                finish_reason = None
+
+            # 从第一个 chunk 获取 role，后续使用这个 role
+            if 'role' in delta and delta['role']:
+                role = delta['role']
+
+            # 获取增量内容
+            delta_content = delta.get('content', '')
+            # reasoning_content = delta.get('reasoning_content', '')
+
+            if remove_think:
+                if delta['content'] is not None:
+                    if '<think>' in delta['content'] and not thinking_started and not thinking_ended:
+                        thinking_started = True
+                        continue
+                    elif delta['content'] == r'</think>' and not thinking_ended:
+                        thinking_ended = True
+                        continue
+                    elif thinking_ended and delta['content'] == '\n\n' and thinking_started:
+                        thinking_started = False
+                        continue
+                    elif thinking_started and not thinking_ended:
+                        continue
+
+            # delta_tool_calls = None
+            if delta.get('tool_calls'):
+                for tool_call in delta['tool_calls']:
+                    if tool_call['id'] and tool_call['function']['name']:
+                        tool_id = tool_call['id']
+                        tool_name = tool_call['function']['name']
+
+                    if tool_call['id'] is None:
+                        tool_call['id'] = tool_id
+                    if tool_call['function']['name'] is None:
+                        tool_call['function']['name'] = tool_name
+                    if tool_call['function']['arguments'] is None:
+                        tool_call['function']['arguments'] = ''
+                    if tool_call['type'] is None:
+                        tool_call['type'] = 'function'
+
+            # 跳过空的第一个 chunk（只有 role 没有内容）
+            if chunk_idx == 0 and not delta_content and not delta.get('tool_calls'):
+                chunk_idx += 1
+                continue
+
+            # 构建 MessageChunk - 只包含增量内容
+            chunk_data = {
+                'role': role,
+                'content': delta_content if delta_content else None,
+                'tool_calls': delta.get('tool_calls'),
+                'is_final': bool(finish_reason),
+            }
+
+            # 移除 None 值
+            chunk_data = {k: v for k, v in chunk_data.items() if v is not None}
+
+            yield provider_message.MessageChunk(**chunk_data)
+            chunk_idx += 1
--- a/src/langbot/pkg/provider/modelmgr/requesters/ppiochatcmpl.yaml
+++ b/src/langbot/pkg/provider/modelmgr/requesters/ppiochatcmpl.yaml
@@ -0,0 +1,39 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: ppio-chat-completions
+  label:
+    en_US: ppio
+    zh_Hans: 派欧云
+  icon: ppio.svg
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://api.ppinfra.com/v3/openai
+  - name: args
+    label:
+      en_US: Args
+      zh_Hans: 附加参数
+    type: object
+    required: true
+    default: {}
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: int
+    required: true
+    default: 120
+  support_type:
+  - llm
+  - text-embedding
+  provider_category: maas
+execution:
+  python:
+    path: ./ppiochatcmpl.py
+    attr: PPIOChatCompletions
--- a/src/langbot/pkg/provider/modelmgr/requesters/qhaigc.png
+++ b/src/langbot/pkg/provider/modelmgr/requesters/qhaigc.png
--- a/src/langbot/pkg/provider/modelmgr/requesters/qhaigcchatcmpl.py
+++ b/src/langbot/pkg/provider/modelmgr/requesters/qhaigcchatcmpl.py
@@ -0,0 +1,17 @@
+from __future__ import annotations
+
+import openai
+import typing
+
+from . import chatcmpl
+
+
+class QHAIGCChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """启航 AI ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://api.qhaigc.com/v1',
+        'timeout': 120,
+    }
--- a/src/langbot/pkg/provider/modelmgr/requesters/qhaigcchatcmpl.yaml
+++ b/src/langbot/pkg/provider/modelmgr/requesters/qhaigcchatcmpl.yaml
@@ -0,0 +1,39 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: qhaigc-chat-completions
+  label:
+    en_US: QH AI
+    zh_Hans: 启航 AI
+  icon: qhaigc.png
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://api.qhaigc.net/v1
+  - name: args
+    label:
+      en_US: Args
+      zh_Hans: 附加参数
+    type: object
+    required: true
+    default: {}
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: int
+    required: true
+    default: 120
+  support_type:
+  - llm
+  - text-embedding
+  provider_category: maas
+execution:
+  python:
+    path: ./qhaigcchatcmpl.py
+    attr: QHAIGCChatCompletions
--- a/src/langbot/pkg/provider/modelmgr/requesters/shengsuanyun.py
+++ b/src/langbot/pkg/provider/modelmgr/requesters/shengsuanyun.py
@@ -0,0 +1,32 @@
+from __future__ import annotations
+
+import openai
+import typing
+
+from . import chatcmpl
+import openai.types.chat.chat_completion as chat_completion
+
+
+class ShengSuanYunChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """胜算云(ModelSpot.AI) ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://router.shengsuanyun.com/api/v1',
+        'timeout': 120,
+    }
+
+    async def _req(
+        self,
+        args: dict,
+        extra_body: dict = {},
+    ) -> chat_completion.ChatCompletion:
+        return await self.client.chat.completions.create(
+            **args,
+            extra_body=extra_body,
+            extra_headers={
+                'HTTP-Referer': 'https://langbot.app',
+                'X-Title': 'LangBot',
+            },
+        )
--- a/src/langbot/pkg/provider/modelmgr/requesters/shengsuanyun.svg
+++ b/src/langbot/pkg/provider/modelmgr/requesters/shengsuanyun.svg
--- a/src/langbot/pkg/provider/modelmgr/requesters/shengsuanyun.yaml
+++ b/src/langbot/pkg/provider/modelmgr/requesters/shengsuanyun.yaml
@@ -0,0 +1,39 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: shengsuanyun-chat-completions
+  label:
+    en_US: ShengSuanYun
+    zh_Hans: 胜算云
+  icon: shengsuanyun.svg
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://router.shengsuanyun.com/api/v1
+  - name: args
+    label:
+      en_US: Args
+      zh_Hans: 附加参数
+    type: object
+    required: true
+    default: {}
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: int
+    required: true
+    default: 120
+  support_type:
+  - llm
+  - text-embedding
+  provider_category: maas
+execution:
+  python:
+    path: ./shengsuanyun.py
+    attr: ShengSuanYunChatCompletions
--- a/src/langbot/pkg/provider/modelmgr/requesters/siliconflow.svg
+++ b/src/langbot/pkg/provider/modelmgr/requesters/siliconflow.svg
@@ -0,0 +1 @@
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" fill="none" version="1.1" width="128" height="128" viewBox="0 0 128 128"><g><g style="opacity:0;"><rect x="0" y="0" width="128" height="128" rx="0" fill="#FFFFFF" fill-opacity="1"/></g><g><path d="M100.74,12L93.2335,12C69.21260000000001,12,55.3672,27.3468,55.3672,50.8672L55.3672,54.8988C52.6011,54.1056,49.7377,53.7031,46.8601,53.7031C29.816499999999998,53.7031,16,67.5196,16,84.5632C16,101.6069,29.816499999999998,115.423,46.8601,115.423C63.9037,115.423,77.72030000000001,101.6069,77.72030000000001,84.5632C77.72030000000001,82.4902,77.51140000000001,80.4223,77.0967,78.3911L77.2197,78.3911L100.74,78.3911C106.9654,78.3681,112,73.3151,112,67.08959999999999C112,60.8642,106.9654,55.8111,100.74,55.7882L100.7362,55.7882L100.6985,55.7879L100.6606,55.7882L77.2197,55.7882L77.2195,49.8663C77.2195,40.8584,83.7252,34.352900000000005,93.2335,34.352900000000005L100.5653,34.352900000000005L100.5733,34.352900000000005L100.5812,34.352900000000005L100.74,34.352900000000005L100.74,34.352900000000005C106.8469,34.2605,111.7497,29.284,111.7497,23.1764C111.7497,17.06889,106.8469,12.0923454,100.74,12L100.74,12ZM56.0347,84.5632C56.0347,79.4962,51.9271,75.3885,46.8601,75.3885C41.793099999999995,75.3885,37.6854,79.4962,37.6854,84.5632C37.6854,89.6303,41.793099999999995,93.7378,46.8601,93.7378C51.9271,93.7378,56.0347,89.6303,56.0347,84.5632Z" fill-rule="evenodd" fill="#8358F6" fill-opacity="1"/></g></g></svg>
--- a/src/langbot/pkg/provider/modelmgr/requesters/siliconflowchatcmpl.py
+++ b/src/langbot/pkg/provider/modelmgr/requesters/siliconflowchatcmpl.py
@@ -0,0 +1,17 @@
+from __future__ import annotations
+
+import typing
+import openai
+
+from . import chatcmpl
+
+
+class SiliconFlowChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """SiliconFlow ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://api.siliconflow.cn/v1',
+        'timeout': 120,
+    }
--- a/src/langbot/pkg/provider/modelmgr/requesters/siliconflowchatcmpl.yaml
+++ b/src/langbot/pkg/provider/modelmgr/requesters/siliconflowchatcmpl.yaml
@@ -0,0 +1,32 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: siliconflow-chat-completions
+  label:
+    en_US: SiliconFlow
+    zh_Hans: 硅基流动
+  icon: siliconflow.svg
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://api.siliconflow.cn/v1
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  - text-embedding
+  provider_category: maas
+execution:
+  python:
+    path: ./siliconflowchatcmpl.py
+    attr: SiliconFlowChatCompletions
--- a/src/langbot/pkg/provider/modelmgr/requesters/tokenpony.svg
+++ b/src/langbot/pkg/provider/modelmgr/requesters/tokenpony.svg
@@ -0,0 +1 @@
+<svg xmlns="http://www.w3.org/2000/svg" width="450" height="280" viewBox="0 0 450 280" class="cursor-pointer h-24 flex-shrink-0 w-149"><g fill="none" fill-rule="nonzero"><path fill="#0005DE" d="M97.705 6.742c58.844 0 90.962 34.353 90.962 98.341v21.843c-15.118-2.479-30.297-6.573-45.558-12.3v-9.543c0-35.97-15.564-56.281-45.404-56.281s-45.404 20.31-45.404 56.281v72.48c0 36.117 15.65 56.818 45.404 56.818 26.78 0 42.133-16.768 44.936-46.452q22.397 6.473 44.905 9.356c-6.15 51.52-37.492 79.155-89.841 79.155-58.678 0-90.963-34.72-90.963-98.878v-72.479c0-63.988 32.119-98.34 90.963-98.34m253.627 0c58.844 0 90.963 34.353 90.963 98.341v72.48c0 64.157-32.285 98.877-90.963 98.877-52.438 0-83.797-27.729-89.874-79.415 15-2.026 29.965-5.252 44.887-9.67 2.658 30.042 18.036 47.026 44.987 47.026 29.755 0 45.404-20.7 45.404-56.819v-72.479c0-35.97-15.564-56.281-45.404-56.281s-45.403 20.31-45.403 56.281v8.778c-15.262 5.868-30.44 10.104-45.559 12.725v-21.503c0-63.988 32.118-98.34 90.962-98.34m-164.37 140.026.57.09.831.127-.83-.128a234.5 234.5 0 0 0 35.979 2.79q18.408.002 36.858-2.928l1.401-.226a242 242 0 0 0 1.45-.244l-1.037.175q.729-.12 1.458-.247l-.421.072 1.26-.219-.84.147a244 244 0 0 0 2.8-.5l-.792.144q.648-.117 1.298-.239l-.506.094q.66-.122 1.322-.248l-.816.154q.759-.142 1.518-.289l-.702.135a247 247 0 0 0 5.364-1.084l-.463.098a250 250 0 0 0 3.928-.864l-.785.178 1.45-.33-.665.152q.597-.137 1.193-.276l-.528.123a253 253 0 0 0 3.685-.882l-.254.063q.683-.168 1.366-.34l-1.112.277q.809-.2 1.618-.405l-.506.128q.818-.206 1.634-.417l-1.128.289q.71-.18 1.419-.365l1.506-.397a259 259 0 0 0 1.804-.488l-.433.119a261 261 0 0 0 3.751-1.053l-.681.196a264 264 0 0 0 1.735-.502l-1.054.306q.636-.184 1.272-.37l-.218.064 1.238-.366-1.02.302a266 266 0 0 0 2.936-.882l-1.026.312q.71-.214 1.42-.433l-.394.121q.675-.207 1.35-.418l-.955.297q.8-.246 1.6-.499l-.645.202q.86-.269 1.72-.543l-1.076.341q.666-.21 1.33-.423l-.254.082q.833-.266 1.665-.539l-1.41.457q.874-.28 1.75-.568l-.34.111q.702-.229 1.403-.462l-1.063.351q.818-.269 1.634-.542l-.571.19a276 276 0 0 0 4.038-1.378l-.735.256q.657-.228 1.315-.46l-.58.204q16.86-5.903 33.78-14.256l-7.114-12.453 42.909 6.553-13.148 45.541-7.734-13.537q-23.832 11.94-47.755 19.504l-.199.063a298 298 0 0 1-11.65 3.412 288 288 0 0 1-10.39 2.603 280 280 0 0 1-11.677 2.431 273 273 0 0 1-11.643 1.903 263.5 263.5 0 0 1-36.858 2.599q-17.437 0-34.844-2.323l-.227-.03q-.635-.085-1.27-.174l1.497.204a268 268 0 0 1-13.673-2.182 275 275 0 0 1-12.817-2.697 282 282 0 0 1-11.859-3.057 291 291 0 0 1-7.21-2.123c-17.23-5.314-34.43-12.334-51.59-21.051l-8.258 14.455-13.148-45.541 42.909-6.553-6.594 11.544q18.421 9.24 36.776 15.572l1.316.45 1.373.462-.831-.278q.795.267 1.589.53l-.758-.252q.632.211 1.264.419l-.506-.167q.642.212 1.284.42l-.778-.253a271 271 0 0 0 3.914 1.251l-.227-.07a267 267 0 0 0 3.428 1.046l-.194-.058 1.315.389-1.121-.331q.864.256 1.73.508l-.609-.177q.826.241 1.651.478l-1.043-.3 1.307.375-.264-.075q.802.228 1.603.452l-1.34-.377q1.034.294 2.067.58l-.727-.203q.713.2 1.426.394l-.699-.192q.62.171 1.237.338l-.538-.146a259 259 0 0 0 3.977 1.051l-.66-.17q.683.177 1.367.35l-.707-.18q.687.175 1.373.348l-.666-.168q.738.186 1.475.368l-.809-.2q.716.179 1.43.353l-.621-.153a253 253 0 0 0 3.766.898l-.308-.07q.735.17 1.472.336l-1.164-.266q.747.173 1.496.34l-.332-.074q.845.19 1.69.374l-1.358-.3q.932.21 1.864.41l-.505-.11q.726.159 1.452.313l-.947-.203q.72.156 1.44.307l-.493-.104q.684.144 1.368.286l-.875-.182q.743.155 1.485.306l-.61-.124q.932.192 1.864.376l-1.254-.252q.904.184 1.809.361l-.555-.109q.752.15 1.504.293l-.95-.184q.69.135 1.377.265l-.427-.081q.784.15 1.569.295l-1.142-.214q.717.136 1.434.268l-.292-.054a244 244 0 0 0 3.808.673l-.68-.116 1.063.18-.383-.064q1.076.18 2.152.352z"></path></g></svg>
--- a/src/langbot/pkg/provider/modelmgr/requesters/tokenpony.yaml
+++ b/src/langbot/pkg/provider/modelmgr/requesters/tokenpony.yaml
@@ -0,0 +1,32 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: tokenpony-chat-completions
+  label:
+    en_US: TokenPony
+    zh_Hans: 小马算力
+  icon: tokenpony.svg
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://api.tokenpony.cn/v1
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  - text-embedding
+  provider_category: maas
+execution:
+  python:
+    path: ./tokenponychatcmpl.py
+    attr: TokenPonyChatCompletions
--- a/src/langbot/pkg/provider/modelmgr/requesters/tokenponychatcmpl.py
+++ b/src/langbot/pkg/provider/modelmgr/requesters/tokenponychatcmpl.py
@@ -0,0 +1,17 @@
+from __future__ import annotations
+
+import typing
+import openai
+
+from . import chatcmpl
+
+
+class TokenPonyChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """TokenPony ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://api.tokenpony.cn/v1',
+        'timeout': 120,
+    }
--- a/src/langbot/pkg/provider/modelmgr/requesters/volcark.svg
+++ b/src/langbot/pkg/provider/modelmgr/requesters/volcark.svg
@@ -0,0 +1,8 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg viewBox="-0.006 0 24.6978 24.9156" xmlns="http://www.w3.org/2000/svg">
+  <path d="M20.511 15.3019L17.2442 28.1928C17.2362 28.2282 17.2364 28.2649 17.2447 28.3001C17.2531 28.3354 17.2694 28.3683 17.2923 28.3964C17.3153 28.4244 17.3444 28.4468 17.3773 28.4619C17.4103 28.477 17.4462 28.4844 17.4825 28.4835H24.0137C24.0499 28.4844 24.0859 28.477 24.1188 28.4619C24.1518 28.4468 24.1809 28.4244 24.2038 28.3964C24.2268 28.3683 24.2431 28.3354 24.2514 28.3001C24.2598 28.2649 24.26 28.2282 24.252 28.1928L20.9685 15.3019C20.9541 15.2524 20.924 15.209 20.8827 15.178C20.8415 15.1471 20.7913 15.1304 20.7397 15.1304C20.6882 15.1304 20.638 15.1471 20.5968 15.178C20.5555 15.209 20.5254 15.2524 20.511 15.3019V15.3019Z" fill="#00E5E5" transform="matrix(1.0178890228271484, 0, 0, 1.0178890228271484, -1.952212187461555e-7, -4.077521402283104)"/>
+  <path d="M2.53051 18.2228L-5.28338e-06 28.1924C-0.00799016 28.2277 -0.00780431 28.2644 0.000538111 28.2997C0.00888053 28.335 0.0251596 28.3679 0.0481365 28.3959C0.0711133 28.4239 0.100182 28.4464 0.133131 28.4615C0.166079 28.4766 0.202039 28.484 0.238273 28.4831H5.28025C5.31649 28.484 5.35245 28.4766 5.38539 28.4615C5.41834 28.4464 5.44741 28.4239 5.47039 28.3959C5.49336 28.3679 5.50964 28.335 5.51799 28.2997C5.52633 28.2644 5.52651 28.2277 5.51853 28.1924L2.98563 18.2228C2.97054 18.1742 2.94032 18.1318 2.89938 18.1016C2.85844 18.0714 2.80892 18.0552 2.75807 18.0552C2.70722 18.0552 2.6577 18.0714 2.61676 18.1016C2.57582 18.1318 2.5456 18.1742 2.53051 18.2228V18.2228Z" fill="#00E5E5" transform="matrix(1.0178890228271484, 0, 0, 1.0178890228271484, -1.952212187461555e-7, -4.077521402283104)"/>
+  <path d="M6.99344 9.96839L2.38275 28.1919C2.37498 28.2263 2.37494 28.262 2.38262 28.2964C2.3903 28.3308 2.40552 28.363 2.42717 28.3908C2.44882 28.4186 2.47637 28.4413 2.50783 28.4572C2.53929 28.473 2.57388 28.4817 2.60911 28.4826H11.8329C11.8691 28.4835 11.9051 28.4761 11.938 28.461C11.971 28.4459 12 28.4235 12.023 28.3955C12.046 28.3675 12.0623 28.3345 12.0706 28.2993C12.079 28.264 12.0791 28.2273 12.0712 28.1919L7.44855 9.96839C7.43347 9.91982 7.40325 9.87736 7.36231 9.8472C7.32136 9.81705 7.27185 9.80078 7.221 9.80078C7.17015 9.80078 7.12063 9.81705 7.07969 9.8472C7.03874 9.87736 7.00852 9.91982 6.99344 9.96839Z" fill="#006EFF" transform="matrix(1.0178890228271484, 0, 0, 1.0178890228271484, -1.952212187461555e-7, -4.077521402283104)"/>
+  <path d="M14.9472 4.17346C14.9321 4.1249 14.9019 4.08244 14.861 4.05228C14.82 4.02213 14.7705 4.00586 14.7197 4.00586C14.6688 4.00586 14.6193 4.02213 14.5784 4.05228C14.5374 4.08244 14.5072 4.1249 14.4921 4.17346L8.18963 28.192C8.18165 28.2273 8.18183 28.264 8.19017 28.2993C8.19852 28.3346 8.2148 28.3675 8.23777 28.3955C8.26075 28.4235 8.28982 28.446 8.32277 28.4611C8.35572 28.4762 8.39168 28.4835 8.42791 28.4827H21.0233C21.0596 28.4835 21.0955 28.4762 21.1285 28.4611C21.1614 28.446 21.1905 28.4235 21.2135 28.3955C21.2364 28.3675 21.2527 28.3346 21.2611 28.2993C21.2694 28.264 21.2696 28.2273 21.2616 28.192L14.9472 4.17346Z" fill="#006EFF" transform="matrix(1.0178890228271484, 0, 0, 1.0178890228271484, -1.952212187461555e-7, -4.077521402283104)"/>
+  <path d="M10.3175 12.6188L6.31915 28.1903C6.31074 28.2258 6.31061 28.2628 6.31875 28.2984C6.3269 28.3339 6.34311 28.3672 6.36614 28.3955C6.38916 28.4238 6.41839 28.4465 6.45155 28.4617C6.48472 28.4769 6.52094 28.4844 6.55743 28.4834H14.535C14.5715 28.4844 14.6077 28.4769 14.6409 28.4617C14.674 28.4465 14.7033 28.4238 14.7263 28.3955C14.7493 28.3672 14.7655 28.3339 14.7737 28.2984C14.7818 28.2628 14.7817 28.2258 14.7733 28.1903L10.7726 12.6188C10.7575 12.5702 10.7273 12.5278 10.6863 12.4976C10.6454 12.4674 10.5959 12.4512 10.545 12.4512C10.4942 12.4512 10.4447 12.4674 10.4037 12.4976C10.3628 12.5278 10.3326 12.5702 10.3175 12.6188Z" fill="#00E5E5" transform="matrix(1.0178890228271484, 0, 0, 1.0178890228271484, -1.952212187461555e-7, -4.077521402283104)"/>
+</svg>
--- a/src/langbot/pkg/provider/modelmgr/requesters/volcarkchatcmpl.py
+++ b/src/langbot/pkg/provider/modelmgr/requesters/volcarkchatcmpl.py
@@ -0,0 +1,17 @@
+from __future__ import annotations
+
+import typing
+import openai
+
+from . import chatcmpl
+
+
+class VolcArkChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """火山方舟大模型平台 ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://ark.cn-beijing.volces.com/api/v3',
+        'timeout': 120,
+    }
--- a/src/langbot/pkg/provider/modelmgr/requesters/volcarkchatcmpl.yaml
+++ b/src/langbot/pkg/provider/modelmgr/requesters/volcarkchatcmpl.yaml
@@ -0,0 +1,31 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: volcark-chat-completions
+  label:
+    en_US: Volc Engine Ark
+    zh_Hans: 火山方舟
+  icon: volcark.svg
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://ark.cn-beijing.volces.com/api/v3
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  provider_category: maas
+execution:
+  python:
+    path: ./volcarkchatcmpl.py
+    attr: VolcArkChatCompletions
--- a/src/langbot/pkg/provider/modelmgr/requesters/xai.svg
+++ b/src/langbot/pkg/provider/modelmgr/requesters/xai.svg
@@ -0,0 +1 @@
+<svg xmlns="http://www.w3.org/2000/svg" fill="currentColor" viewBox="0 0 24 24" aria-hidden="true" class="" focusable="false" style="fill:currentColor;height:28px;width:28px"><path d="m3.005 8.858 8.783 12.544h3.904L6.908 8.858zM6.905 15.825 3 21.402h3.907l1.951-2.788zM16.585 2l-6.75 9.64 1.953 2.79L20.492 2zM17.292 7.965v13.437h3.2V3.395z"></path></svg>
--- a/src/langbot/pkg/provider/modelmgr/requesters/xaichatcmpl.py
+++ b/src/langbot/pkg/provider/modelmgr/requesters/xaichatcmpl.py
@@ -0,0 +1,17 @@
+from __future__ import annotations
+
+import typing
+import openai
+
+from . import chatcmpl
+
+
+class XaiChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """xAI ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://api.x.ai/v1',
+        'timeout': 120,
+    }
--- a/src/langbot/pkg/provider/modelmgr/requesters/xaichatcmpl.yaml
+++ b/src/langbot/pkg/provider/modelmgr/requesters/xaichatcmpl.yaml
@@ -0,0 +1,31 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: xai-chat-completions
+  label:
+    en_US: xAI
+    zh_Hans: xAI
+  icon: xai.svg
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://api.x.ai/v1
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  provider_category: manufacturer
+execution:
+  python:
+    path: ./xaichatcmpl.py
+    attr: XaiChatCompletions
--- a/src/langbot/pkg/provider/modelmgr/requesters/zhipuai.svg
+++ b/src/langbot/pkg/provider/modelmgr/requesters/zhipuai.svg
--- a/src/langbot/pkg/provider/modelmgr/requesters/zhipuaichatcmpl.py
+++ b/src/langbot/pkg/provider/modelmgr/requesters/zhipuaichatcmpl.py
@@ -0,0 +1,17 @@
+from __future__ import annotations
+
+import typing
+import openai
+
+from . import chatcmpl
+
+
+class ZhipuAIChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """智谱AI ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://open.bigmodel.cn/api/paas/v4',
+        'timeout': 120,
+    }
--- a/src/langbot/pkg/provider/modelmgr/requesters/zhipuaichatcmpl.yaml
+++ b/src/langbot/pkg/provider/modelmgr/requesters/zhipuaichatcmpl.yaml
@@ -0,0 +1,31 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: zhipuai-chat-completions
+  label:
+    en_US: ZhipuAI
+    zh_Hans: 智谱 AI
+  icon: zhipuai.svg
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://open.bigmodel.cn/api/paas/v4
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  provider_category: manufacturer
+execution:
+  python:
+    path: ./zhipuaichatcmpl.py
+    attr: ZhipuAIChatCompletions
--- a/src/langbot/pkg/provider/modelmgr/token.py
+++ b/src/langbot/pkg/provider/modelmgr/token.py
@@ -0,0 +1,24 @@
+from __future__ import annotations
+
+import typing
+
+
+class TokenManager:
+    """鉴权 Token 管理器"""
+
+    name: str
+
+    tokens: list[str]
+
+    using_token_index: typing.Optional[int] = 0
+
+    def __init__(self, name: str, tokens: list[str]):
+        self.name = name
+        self.tokens = tokens
+        self.using_token_index = 0
+
+    def get_token(self) -> str:
+        return self.tokens[self.using_token_index]
+
+    def next_token(self):
+        self.using_token_index = (self.using_token_index + 1) % len(self.tokens)
				`@@ -0,0 +1 @@`
				<svg height="1em" style="flex:none;line-height:1" viewBox="0 0 24 24" width="1em" xmlns="http://www.w3.org/2000/svg"><title>Gemini</title><defs><linearGradient id="lobe-icons-gemini-fill" x1="0%" x2="68.73%" y1="100%" y2="30.395%"><stop offset="0%" stop-color="#1C7DFF"></stop><stop offset="52.021%" stop-color="#1C69FF"></stop><stop offset="100%" stop-color="#F0DCD6"></stop></linearGradient></defs><path d="M12 24A14.304 14.304 0 000 12 14.304 14.304 0 0012 0a14.305 14.305 0 0012 12 14.305 14.305 0 00-12 12" fill="url(#lobe-icons-gemini-fill)" fill-rule="nonzero"></path></svg>
				`@@ -0,0 +1 @@`
				<svg id="_层_2" xmlns="http://www.w3.org/2000/svg" viewBox="0 0 274.37 172.76"><defs><style>.cls-2{fill:#36cfd1}.cls-3{fill:#624aff}</style></defs><g id="_层_1-2"><path class="cls-3" d="M24.78 73.55h25.65V99.2H24.78zm99.14 25.66h25.65v25.65h-25.65zm76.95 25.65h-25.65v22.19h47.84V99.21h-22.19v25.65z"/><path class="cls-2" d="M149.57 73.55h25.65V99.2h-25.65zM24.78 47.9h25.65v25.65H24.78z"/><path class="cls-3" d="M223.06 73.55h25.65V99.2h-25.65z"/><path class="cls-2" d="M223.06 47.9h25.65v25.65h-25.65z"/><path class="cls-3" d="M175.22 25.71V47.9h25.65v25.65h22.19V25.71h-47.84z"/><path class="cls-2" d="M98.27 73.55h25.65V99.2H98.27z"/><path class="cls-3" d="M72.62 47.9h25.65V25.71H50.43v47.84h22.19V47.9zm0 51.31H50.43v47.84h47.84v-22.19H72.62V99.21z"/><path style="fill:none" d="M0 0h274.37v172.76H0z"/></g></svg>
				`@@ -0,0 +1 @@`
				<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" fill="none" version="1.1" width="128" height="128" viewBox="0 0 128 128"><g><g style="opacity:0;"><rect x="0" y="0" width="128" height="128" rx="0" fill="#FFFFFF" fill-opacity="1"/></g><g><path d="M100.74,12L93.2335,12C69.21260000000001,12,55.3672,27.3468,55.3672,50.8672L55.3672,54.8988C52.6011,54.1056,49.7377,53.7031,46.8601,53.7031C29.816499999999998,53.7031,16,67.5196,16,84.5632C16,101.6069,29.816499999999998,115.423,46.8601,115.423C63.9037,115.423,77.72030000000001,101.6069,77.72030000000001,84.5632C77.72030000000001,82.4902,77.51140000000001,80.4223,77.0967,78.3911L77.2197,78.3911L100.74,78.3911C106.9654,78.3681,112,73.3151,112,67.08959999999999C112,60.8642,106.9654,55.8111,100.74,55.7882L100.7362,55.7882L100.6985,55.7879L100.6606,55.7882L77.2197,55.7882L77.2195,49.8663C77.2195,40.8584,83.7252,34.352900000000005,93.2335,34.352900000000005L100.5653,34.352900000000005L100.5733,34.352900000000005L100.5812,34.352900000000005L100.74,34.352900000000005L100.74,34.352900000000005C106.8469,34.2605,111.7497,29.284,111.7497,23.1764C111.7497,17.06889,106.8469,12.0923454,100.74,12L100.74,12ZM56.0347,84.5632C56.0347,79.4962,51.9271,75.3885,46.8601,75.3885C41.793099999999995,75.3885,37.6854,79.4962,37.6854,84.5632C37.6854,89.6303,41.793099999999995,93.7378,46.8601,93.7378C51.9271,93.7378,56.0347,89.6303,56.0347,84.5632Z" fill-rule="evenodd" fill="#8358F6" fill-opacity="1"/></g></g></svg>
				`@@ -0,0 +1 @@`
				`<svg xmlns="http://www.w3.org/2000/svg" fill="currentColor" viewBox="0 0 24 24" aria-hidden="true" class="" focusable="false" style="fill:currentColor;height:28px;width:28px"><path d="m3.005 8.858 8.783 12.544h3.904L6.908 8.858zM6.905 15.825 3 21.402h3.907l1.951-2.788zM16.585 2l-6.75 9.64 1.953 2.79L20.492 2zM17.292 7.965v13.437h3.2V3.395z"></path></svg>`