LangChain
← Back to Overview: Agents
What Is LangChain
LangChain is a framework for building LLM-powered applications using composable chains and agents. It provides abstractions for memory, tools, retrieval, and model integrations.
When to Use LangChain
- Building retrieval-augmented generation (RAG) pipelines
- Simple to moderate agent complexity with standard tool use
- You need broad model and tool integrations out of the box
LangChain vs LangGraph
LangGraph is built on top of LangChain and adds graph-based state machines for more complex, cyclical agent workflows. For simple agents, LangChain is sufficient; for stateful/long-horizon agents, prefer LangGraph.
Chapter Map
| File | Topic |
|---|---|
| 01 — LangChain Fundamentals | Core abstractions, setup |
| 02 — Simple Agent | Basic agent with tools |
| 03 — Complex Agent | Multi-tool, memory-enabled agent |