Light
Dark
left click to zoom
LLM Agents
LLM Agents
LLM Agents
LLM Agents
LLM Agents

The Operating System for AI Agents

Intelligent agents that remember, learn, and scale seamlessly

Install LettaRequest Cloud Access
deployment options

Run Locally

View Docs

Use the Agent Development Environment (ADE) to visualize your agent's memory, reasoning steps, and tool calls. Understand decision making and make edits to your agent's state in real time.

Letta Cloud

Get Started

Take your agents from prototype to production with our fully managed cloud offering. Connect via API keys or use the cloud ADE for a seamless development experience.

trusted by
agent developers
PLAY

Your Agents Need an OS: Today's AI agents struggle with fragmented context and limited memory. Just as computers needed operating systems to manage resources, your agents need intelligent context management to unlock their full potential.

Iterate on your agents at the speed of thought

With the Agent Development Environment (ADE), you can visualize your agent’s memory, reasoning steps, and tool calls to understand their decision making, and make edits to their state in real time.

Try the ADE today

Built by the researchers behind MemGPT

Maintain memory across sessions for long running conversations. Ensure your agents can learn from interactions and improve over time, rather than derail.

Read the MemGPT paper

No more context window overflow

Overstuffed context windows confuse agents and degrade reasoning. Letta maximizes agent performance by compiling the most relevant information to pass to the LLM, while keeping token counts under a specified budget.

Learn about the Letta platform

Deploy and monitor agents using APIs

Use our REST API or Python and Node.js (TypeScript) SDKs to create agents running on a Letta server. Directly query or modify agent state, configurations, and message histories.

Browse our API Documentation

Scale your application to millions of agents

Deploy a Letta service on Kubernetes using Docker. Monitor and query real-time agent state persisted in Postgres. Execute agent tools in isolated sandboxes with configurable environment variables and secrets.

Deploy a Letta Server
GitHub
Contribute to the Letta open source
Visit our GitHub
Documentation
Learn how to use the Letta AI platform
Read our docs
Discord
Join our developer Discord community
Join our Discord

About the company

Born out of the Sky Computing Lab at UC Berkeley, Letta believes that programming agents starts with programming memory.

Read more

We're hiring! Will you join us?

Build the next generation of agentic systems
Our values
Open AI and model freedom
Power to the developers
Nimble, talent-dense teams
Flat structure & high trust
We
in-person work
We offer
Top tier medical coverage
Competitive salary & equity
Workstation stipend
Office located in the heart of SF
Additional benefits
See open positions

Latest news

Letta
Apr 21, 2025

Sleep-time Compute

Sleep-time compute is a new way to scale AI capabilities: letting models "think" during downtime. Instead of sitting idle between tasks, AI agents can now use their "sleep" time to process information and form new connections by rewriting their memory state.

Letta
Apr 17, 2025

Announcing Letta Client SDKs for Python and TypeScript

We've releasing new client SDKs (support for TypeScript and Python) and upgraded developer documentation

Letta
GitHub
Apr 2, 2025

Agent File

Introducing Agent File (.af): An open file format for serializing stateful agents with persistent memory and behavior.

Around the web

Press
Sleep Time Compute - AI That "Thinks" 24/7 (Breakthrough)

What if AI models could figure out answers to your questions BEFORE you even ask them? An incredible new research paper introduces "Sleep-Time Compute," allowing models to think while they aren't being used.

Press
MemGPT with Real-life Example: Bridging the Gap Between AI and OS

In this tutorial, we discuss and show how to run MemGPT - an LLM with the potential for infinite context understanding.

Press
MemGPT: Unlimited Memory without Token Constraints for Generation

MemGPT, standing for Memory-GPT, is a system devised to enhance the performance of Large Language Models (LLMs) by introducing a more advanced memory management scheme, helping to overcome the challenges posed by fixed context windows. Below are some of the key features of MemGPT: