Light
Dark
LLM Agents
LLM Agents
LLM Agents
LLM Agents
LLM Agents

Making machines that learn

We’re solving AI’s memory problem. Create agents that remember everything, learn continuously, and improve themselves over time.

Letta PlatformLetta Code
Build

Letta Platform

Read More

Give your agents the ability to remember and learn, so they can be more like humans, and less like bots.

Explore our tools
Letta Code
A memory-first, model agnostic coding agent
Letta API
A low-level API for managing your agent's memory and context
Real agents, in production
Discover

AI Research

Read More

We are an open AI lab pursuing foundational research in AI memory, continual learning, and self-improvement.

Read our research
Continual Learning in Token Space
Context-Bench: Evaluating Memory
Sleep-time Compute: Learning Offline
MemGPT: The LLM Operating System
View all research

Today's AI agents struggle to remember previous mistakes, and are unable to learn from new experiences. At Letta, we're building machines with real memory that can continually learn and self-improve.

digital humans
PLAY
Letta is an open AI lab building intelligent machines that can learn from experience.
About usOpen roles

Latest articles

Read more
Research
Feb 12, 2026

Introducing Context Repositories: Git-based Memory for Coding Agents

We're introducing Context Repositories, a rebuild of how memory works in Letta Code based on programmatic context management and git-based versioning.

Product
Jan 21, 2026

Conversations: Shared Agent Memory across Concurrent Experiences

The Conversations API allows you to build agents that can maintain shared memory across parallel experiences with users

Product
Dec 16, 2025

Letta Code: A Memory-First Coding Agent

Introducing Letta Code, a memory-first coding agent. Letta Code is the #1 model-agnostic open source agent on the leading AI coding benchmark Terminal-Bench.

Around the web

August 20, 2025
Do Large Language Models Dream of AI Agents?

For AI models, knowing what to remember might be as important as knowing what to forget. Welcome to the era of “sleep-time compute.”

July 16, 2025
Why sleep-time compute is the next big leap in AI

Letta’s model-agnostic framework lets AI agents think while idle, combining memory refinement, precomputed responses, and multi-model setups to build smarter, always-on systems.

April 25, 2025
Sleep Time Compute - AI That "Thinks" 24/7

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.