letta
Platform for stateful agents: AI with advanced memory that can learn and self-improve over time.
What is letta?
Platform for stateful agents: AI with advanced memory that can learn and self-improve over time.
Overview of letta
What it is
Letta is a platform for building stateful agents with advanced memory. It allows AI agents to learn and self-improve over time. Formerly known as MemGPT, Letta includes an agent harness, interactive terminal UI, App Server, channels, and the runtime used by desktop and web applications. The source code is available under the Apache-2.0 license.
Key features
- Stateful agents with memory that can learn and improve over time - Interactive terminal UI for direct agent interaction - App Server for local or self-hosted agent deployment - Support for multiple channels: Slack, Telegram, Discord, and custom channels - Letta Agent SDK for building agents into TypeScript applications - Letta Cloud for keeping agent memory, identity, and conversations available across computers - Open source (Apache-2.0)
Use cases
- Building AI agents that require persistent memory and self-improvement - Deploying agents on chat platforms (Slack, Telegram, Discord, custom channels)
- Using cloud infrastructure to maintain agent state across different computers - Integrating stateful agents into TypeScript applications via the SDK
Getting started
Install Letta from npm: bash npm install -g @letta-ai/letta-code Launch the interactive terminal UI: bash letta Run the App Server for local or self-hosted agents: bash letta server See the Letta documentation for current installation, development, and deployment instructions.
License and pricing
The license is Apache-2.0. The project is open source with no stated pricing.
Based on the project's public README and repository metadata.