Reference

Welcome to the LangGraph reference docs! These pages detail the core interfaces you will use when building with LangGraph. Each section covers a different part of the ecosystem.

Tip

If you are just getting started, see LangGraph basics for an introduction to the main concepts and usage patterns.

LangGraph

The core APIs for the LangGraph opens source library.

• Graphs: Main graph abstraction and usage.
• Functional API: Functional programming interface for graphs.
• Pregel: Pregel-inspired computation model.
• Checkpointing: Saving and restoring graph state.
• Storage: Storage backends and options.
• Caching: Caching mechanisms for performance.
• Types: Type definitions for graph components.
• Config: Configuration options.
• Errors: Error types and handling.
• Constants: Global constants.
• Channels: Message passing and channels.

Prebuilt components

Higher-level abstractions for common workflows, agents, and other patterns.

• Agents: Built-in agent patterns.
• Supervisor: Orchestration and delegation.
• Swarm: Multi-agent collaboration.
• MCP Adapters: Integrations with external systems.

LangGraph Platform

Tools for deploying and connecting to the LangGraph Platform.

• CLI: Command-line interface for building and deploying LangGraph Platform applications.
• Server API: REST API for the LangGraph Server.
• SDK (Python): Python SDK for interacting with instances of the LangGraph Server.
• SDK (JS/TS): JavaScript/TypeScript SDK for interacting with instances of the LangGraph Server.
• RemoteGraph: Pregel abstraction for connecting to LangGraph Server instances.
• Environment variables: Supported configuration variables when deploying with the LangGraph Platform.