Letta Documentation
Create stateful AI agents that truly remember, learn, and evolve.
Letta enables you to build and deploy stateful AI agents that maintain memory and context across long-running conversations. Develop agents that truly learn and evolve from interactions without starting from scratch each time.
Build agents with intelligent memory, not limited context
Letta’s advanced context management system - built by the researchers behind MemGPT - transforms how agents remember and learn. Unlike basic agents that forget when their context window fills up, Letta agents maintain memories across sessions and continuously improve, even while they sleep .
Start building in minutes
Our quickstart and examples work on both Letta Cloud and self-hosted Letta.
Create your first stateful agent using the Letta API & ADE
Build a full agents application using create-letta-app
Build stateful agents with your favorite tools
Connect to agents running in a Letta server using any of your preferred development frameworks. Letta integrates seamlessly with the developer tools you already know and love.
Core SDK for our REST API
Core SDK for our REST API
Framework integration
Framework integration
Framework integration
Framework integration
See what your agents are thinking
The Agent Development Environment (ADE) provides complete visibility into your agent’s memory, context window, and decision-making process - essential for developing and debugging production agent applications.
Run agents as services, not libraries
Letta is fundamentally different from other agent frameworks. While most frameworks are libraries that wrap model APIs, Letta provides a dedicated service where agents live and operate autonomously. Agents continue to exist and maintain state even when your application isn’t running, with computation happening on the server and all memory, context, and tool connections handled by the Letta server.
Everything you need for production agents
Letta provides a complete suite of capabilities for building and deploying advanced AI agents:
- Agent Development Environment (agent builder + monitoring UI)
- Python SDK + TypeScript SDK + REST API
- Memory management
- Persistence (all agent state is stored in a database)
- Tool calling & execution (support for custom tools & pre-made tools)
- Tool rules (constraining an agent’s action set in a graph-like structure)
- Streaming support
- Native multi-agent support and multi-user support
- Model-agnostic across closed (OpenAI, etc.) and open providers (LM Studio, vLLM, etc.)
- Production-ready deployment (self-hosted with Docker or Letta Cloud)
Join our developer community
Building something with Letta? Join our Discord to connect with other developers creating stateful agents and share what you’re working on.