LangChain
The foundational framework for building with LLMs
What it is
LangChain is the starting point for anyone building LLM-powered apps. It provides abstractions for prompt chains, vector database integration, memory management, and external tool calls. Massive community and thousands of ready integrations. Great starting point; migrate to LangGraph for more complex cases.
How to use it, step by step
- 1
Install the library: pip install langchain langchain-openai (or your preferred provider).
- 2
Set your model's API key as an environment variable.
- 3
Create a model with ChatOpenAI() and send a first message to confirm it works.
- 4
Build a 'chain' by connecting a PromptTemplate to the model with the | (pipe) operator, then run it with .invoke().
- 5
Add memory, document retrieval, or tool calls as your app grows.
💡Practical tips
- →Start from the official docs: the 'getting started' examples cover 90% of basic cases.
- →For RAG (answers over your documents) use a vector store like Chroma or Pinecone with LangChain retrievers.
- →When the app gets complex with loops and state, consider moving to LangGraph.
💰Pricing
Free and open-source framework. You only pay for the models and external services you connect.
The official site opens in a new tab.