The Open-Source AI Stack
RSS

The Stack · Agents · Open source

LangChain

Most-used open agent framework; LLM tools, chains, RAG primitives, and LangGraph for low-level agent orchestration. MCP via the langchain-mcp-adapters package.

LangChain is the most-used open framework for building LLM applications: tool integrations, chains, RAG primitives, and, through the LangGraph spinoff, low-level graph-based agent orchestration. LangChain 1.0 and LangGraph 1.0 both reached general availability in October 2025. It supports Python and JavaScript/TypeScript. Against the opinionated, batteries-included style of CrewAI, LangGraph sits at the low-level end: explicit state machines and graphs the developer wires together. Two facts matter for a 2026 comparison. MCP support is not in the core library; it is provided through the separate langchain-mcp-adapters package, which makes LangChain a later adopter of MCP than the vendor SDKs. License: the core libraries are MIT, but the langgraph-api server runtime used for hosted deployment is under the Elastic License 2.0, which needs a commercial key in production. For open-source AI, LangChain is fully model-agnostic and the default starting point for many teams. LangSmith, the observability and evaluation product, is a separate hosted service that the framework does not require.

Sources

Want a follow-up? Ask the chat about LangChain in context. It will compare to siblings at the same layer and ground every claim in the wiki.

Other projects at the Agents layer

19 siblings · ordered open first