🎉 学 LangChain 免费领 openAI GPT key 限额1000名 →

LangChain 🦜️🔗 中文网，跟着LangChain一起学LLM/GPT开发 JS/TS LangChainJS/TS LangChain (opens in a new tab)Python LangChainPython LangChain (opens in a new tab) OpenAI 中文文档 OpenAI 中文文档 (opens in a new tab)

GitHub (opens in a new tab)

开始
- 索引（Index）
- 演练（Walkthrough）
包版本说明（Packages)
安全（Security)
生态（Ecosystem)
参考资料（Reference)
- Memory
  Build a custom RAG agent with LangGraph
  Application structure
  Backward compatibility
  Case studies
  Checkpointers
  Choosing between Graph and Functional APIs
  Deployment
  Event streaming
  Fault tolerance
  Functional API overview
  Graph API overview
  Install LangGraph
  Interrupts
  Run a local server
  LangSmith Observability
  overview
  Persistence
  runtime
  Quickstart
  Build a custom SQL agent
  Stores
  Streaming
  LangSmith Studio
  Test
  Thinking in LangGraph
  Agent Chat UI
  Use the functional API
  Use the graph API
  Subgraphs
  Use time-travel
  Workflows and agents
  GRAPH_RECURSION_LIMIT
  INVALID_CHAT_HISTORY
  INVALID_CONCURRENT_GRAPH_UPDATE
  INVALID_GRAPH_NODE_RETURN_VALUE
  MISSING_CHECKPOINTER
  MULTIPLE_SUBGRAPHS
  Custom stream channels
  Graph execution
  Overview
- LangChain Academy
  Agents
  Component architecture
  Context engineering in agents
  Build a data analysis agent from scratch
  Deployment
  Event streaming
  Get help
  Guardrails
  Human-in-the-loop
  Install LangChain
  Build a semantic search engine with LangChain
  Long-term memory
  Messages
  LangSmith Observability
  Philosophy
  Retrieval
  Runtime
  Short-term memory
  Build a SQL agent
  Streaming
  Structured output
  LangSmith Studio
  Tools
  Agent Chat UI
  Build a voice agent with LangChain
  Models
  Overview
  Quickstart
  INVALID_PROMPT_INPUT
  INVALID_TOOL_RESULTS
  MESSAGE_COERCION_FAILURE
  MODEL_AUTHENTICATION
  MODEL_NOT_FOUND
  MODEL_RATE_LIMIT
  OUTPUT_PARSING_FAILURE
  Branching chat
  Controlled generative UI
  Declarative generative UI
  Generative UI overview
  Headless tools
  Human-in-the-Loop
  Join & rejoin streams
  Markdown messages
  Message queues
  Open-ended generative UI
  Reasoning tokens
  Structured output
  Time travel
  Tool calling
  AI Elements
  assistant-ui
  CopilotKit
  OpenUI
  Authentication
  Connections
  Model Context Protocol (MCP)
  Tools
  Prebuilt middleware
  Custom middleware
  Custom workflow
  Build customer support with handoffs
  Handoffs
  Multi-agent
  Build a multi-source knowledge base with routing
  Router
  Build a SQL assistant with on-demand skills
  Skills
  Build a personal assistant with subagents
  Subagents
  Agent Evals
  Test
  Integration testing
  Unit testing

开始
- 索引（Index）
- 演练（Walkthrough）
包版本说明（Packages)
安全（Security)
生态（Ecosystem)
参考资料（Reference)
JS/TS LangChain (opens in a new tab)
Python LangChain (opens in a new tab)
OpenAI 中文文档 (opens in a new tab)
- Pinecone 中文文档 (opens in a new tab)
- Milvus 中文文档 (opens in a new tab)
- Memory
  Build a custom RAG agent with LangGraph
  Application structure
  Backward compatibility
  Case studies
  Checkpointers
  Choosing between Graph and Functional APIs
  Deployment
  Event streaming
  Fault tolerance
  Functional API overview
  Graph API overview
  Install LangGraph
  Interrupts
  Run a local server
  LangSmith Observability
  overview
  Persistence
  runtime
  Quickstart
  Build a custom SQL agent
  Stores
  Streaming
  LangSmith Studio
  Test
  Thinking in LangGraph
  Agent Chat UI
  Use the functional API
  Use the graph API
  Subgraphs
  Use time-travel
  Workflows and agents
  GRAPH_RECURSION_LIMIT
  INVALID_CHAT_HISTORY
  INVALID_CONCURRENT_GRAPH_UPDATE
  INVALID_GRAPH_NODE_RETURN_VALUE
  MISSING_CHECKPOINTER
  MULTIPLE_SUBGRAPHS
  Custom stream channels
  Graph execution
  Overview
- LangChain Academy
  Agents
  Component architecture
  Context engineering in agents
  Build a data analysis agent from scratch
  Deployment
  Event streaming
  Get help
  Guardrails
  Human-in-the-loop
  Install LangChain
  Build a semantic search engine with LangChain
  Long-term memory
  Messages
  LangSmith Observability
  Philosophy
  Retrieval
  Runtime
  Short-term memory
  Build a SQL agent
  Streaming
  Structured output
  LangSmith Studio
  Tools
  Agent Chat UI
  Build a voice agent with LangChain
  Models
  Overview
  Quickstart
  INVALID_PROMPT_INPUT
  INVALID_TOOL_RESULTS
  MESSAGE_COERCION_FAILURE
  MODEL_AUTHENTICATION
  MODEL_NOT_FOUND
  MODEL_RATE_LIMIT
  OUTPUT_PARSING_FAILURE
  Branching chat
  Controlled generative UI
  Declarative generative UI
  Generative UI overview
  Headless tools
  Human-in-the-Loop
  Join & rejoin streams
  Markdown messages
  Message queues
  Open-ended generative UI
  Reasoning tokens
  Structured output
  Time travel
  Tool calling
  AI Elements
  assistant-ui
  CopilotKit
  OpenUI
  Authentication
  Connections
  Model Context Protocol (MCP)
  Tools
  Prebuilt middleware
  Custom middleware
  Custom workflow
  Build customer support with handoffs
  Handoffs
  Multi-agent
  Build a multi-source knowledge base with routing
  Router
  Build a SQL assistant with on-demand skills
  Skills
  Build a personal assistant with subagents
  Subagents
  Agent Evals
  Test
  Integration testing
  Unit testing

Question? Give us feedback → (opens in a new tab)Edit this page

新版文档(v1.x)

langchain

LangSmith Studio

LangSmith Studio

import StudioPy from '/snippets/oss/studio-py.mdx'; import StudioJs from '/snippets/oss/studio-js.mdx';

For more information about deployed agents, see Deployment (opens in a new tab).

Structured output Tools

MIT 2026 © Langchain中文网. 跟着langchain学AI应用开发 GitHub | GraphRAG中文网 | chatgpt中文网 |

Langchain英文站 | Langchain JS/TS 文档 | Favicon.ID 工具 |微热文 - 今日热榜 | Model Context Protocol