Python agent framework
LangChain
A framework for building agents and LLM powered applications with interoperable components and third party integrations.
Quick decision
What it is
LangChain is described in its repository as a framework for building agents and LLM powered applications that helps chain together interoperable components and third party integrations. The PyPI page positions it as a quick way to connect models such as OpenAI, Anthropic and Google through a prebuilt agent architecture.
The project emphasizes model interoperability, real time data augmentation through a large integration library, and rapid prototyping with modular components. It also points users with more advanced orchestration needs toward LangGraph, noting that LangChain agents are built on LangGraph for durable execution, streaming, human in the loop and persistence.
LangChain is part of a wider ecosystem that includes LangGraph for low level orchestration, LangSmith for evaluation and observability, and deployment tooling for long running stateful workflows. The Python package is the primary record here, with an equivalent JavaScript and TypeScript library referenced from the repository.
What it does
- Provides a standard interface for chat models, embedding models, vector stores, retrievers and tools.
- Ships a prebuilt agent architecture built on LangGraph with durable execution, streaming and persistence.
- Offers a large official integrations catalog covering search, scraping, browser and platform tools.
- Supports swapping model providers through common abstractions for experimentation.
- Connects to LangSmith for tracing, evaluation, observability and debugging.
What it does not do
- It is not itself a hosted search, scraping or browser automation service; those capabilities come from integration partners and their API keys.
- It is not the low level orchestration runtime for heavily custom deterministic workflows; the docs direct those users to LangGraph.
- It does not remove the need to configure provider credentials, tool permissions and production monitoring separately.
Setup
Install LangChain and call a model
The repository quickstart installs the Python package and initializes a chat model through a provider string.
uv add langchain
from langchain.chat_models import init_chat_model
model = init_chat_model("openai:gpt-5.5")
result = model.invoke("Hello, world!")Environment variables (names only)
OPENAI_API_KEY
Source: https://github.com/langchain-ai/langchain. Examples use placeholders only.
MCP support: LangChain documents MCP tool use through the langchain-mcp-adapters package, which makes MCP server tools compatible with LangChain and LangGraph agents.
Health and maintenance
147,543
Checked 2026-10-07
24,725
Checked 2026-10-07
Stars and forks read from the GitHub repository page in this run. PyPI page shows release files for langchain 1.4.3.
Verified compatible tools
Via langchain-tavily (TavilySearch, TavilyExtract, TavilyCrawl, TavilyMap). The LangChain Tavily provider page states the langchain-tavily package exposes Tavily Search, Extract, Crawl and Map endpoints as LangChain tools and requires TAVILY_API_KEY. Evidence. Setup detail
Via langchain-exa (ExaSearchResults, ExaFindSimilarResults, ExaSearchRetriever). The LangChain Exa provider page states the Exa integration exists in its own partner package, requires EXA_API_KEY, and provides a retriever and agent tools. Evidence. Setup detail
Via langchain-firecrawl (FirecrawlScrape, FirecrawlCrawl, FirecrawlMap, FirecrawlExtract, FirecrawlSearch). The LangChain tool integrations index lists Firecrawl Search in its search tools table, and a fetched LangChain docs pull request describes the langchain-firecrawl package and its five tools. Evidence. Setup detail
Via LangChain Playwright tool integration listed in the Python tool integrations index. The fetched LangChain Python tool integrations listing includes Playwright among the available tool integrations for LangChain. Evidence. Setup detail
Via Composio integration listed in the LangChain tool integrations index. The LangChain tool integrations index lists Composio under integration platforms with 500 plus integrations, OAuth handling and event driven workflows, and also lists Composio in all tools. Evidence. Setup detail
Via LangSmith (same vendor LangChain ecosystem). The LangChain repository README lists LangSmith for agent evals, observability and debugging, and for developing, debugging and deploying AI agents and LLM applications. Evidence. Setup detail
Grouped view: tools for LangChain.
Sources and freshness
- GitHub repository: https://github.com/langchain-ai/langchain
- PyPI package: https://pypi.org/project/langchain/
- Tool integrations index: https://docs.langchain.com/oss/python/integrations/tools
Last checked 2026-10-07. Verification label: source verified. It does not mean AgentsUse ran the framework. Spotted an error? Send a correction. Back to the framework index.