Letta
Open-source framework for building stateful LLM agents with persistent memory - formerly MemGPT - with 15K+ GitHub stars and a managed cloud platform.
Letta (formerly MemGPT) is an open-source framework from UC Berkeley researchers for building LLM agents that maintain persistent, long-term memory beyond the context window. By managing in-context and out-of-context memory storage like an operating system manages RAM and disk, Letta enables agents that remember every past interaction indefinitely. The platform offers both a self-hosted OSS option (Apache 2.0) and Letta Cloud for managed deployment. Backed by $10M seed funding from a16z, Letta has become a core building block for production AI agents requiring stateful context - with 15K+ GitHub stars and active enterprise adoption.
Key Features
- Persistent agent memory with in-context and out-of-context storage management beyond the context window
- Stateful agents that remember full conversation history across sessions indefinitely
- Tool and function calling with built-in memory read/write capabilities per agent
- Multi-agent orchestration with shared memory pools across agent teams
- REST API and Python SDK for building and deploying production memory-enabled agents
- Letta Cloud for managed agent hosting with observability and analytics dashboards
- Apache 2.0 open-source license with self-hosting on any infrastructure
Use Cases
- Building customer support bots that retain full context of every past customer interaction without summarization loss
- Creating personal AI assistants that accumulate a persistent model of user preferences and working style
- Developing long-running research agents that continuously refine and expand a knowledge base
- Enterprise AI workflows requiring stateful, context-aware automation across multi-step business processes
Pros
- Persistent memory solves the context window limitation that makes most LLM agents forget between sessions
- Open-source with active development - 15K+ GitHub stars and backed by a16z for long-term roadmap confidence
- Multi-agent shared memory enables complex team workflows that single-agent frameworks cannot support
Cons
- Steeper learning curve than simpler agent frameworks like CrewAI - memory architecture requires upfront design
- Letta Cloud pricing scales with usage and can become expensive at high agent-call volumes in production
- Memory management concepts require architectural planning to implement correctly from the start
Letta Alternatives
Explore similar tools and alternatives
Looking for alternatives to Letta? Here are some similar tools you might like:
CrewAI
Open-source multi-agent AI framework - 450M+ workflows/month, 60% of Fortune 500 use it; free tier with 50 executions/month, $44.5M raised.
Phidata
Open-source Python framework for building multi-modal AI agents with memory, knowledge, and tool use - 16k+ GitHub stars.
AgentGPT
Browser-based autonomous AI agent platform that lets users deploy goal-directed AI agents without coding, with 30k+ GitHub stars and free tier access.
Ready to try Letta?
Visit the official website to explore all features and get started with Letta today.
Reviews
0 reviews for Letta
Based on 0 reviews
Share your experience
Log in to write a review for Letta
Ito
Only code review that runs your code. Provides runtime analysis with evidence (logs, video, screenshot) to show how code changes application actually work. Back-end, front-end, api, integration.
Cursor
AI-native code editor built on VS Code with built-in AI chat, autocomplete, and codebase understanding.
GitHub Copilot
AI pair programmer by GitHub/OpenAI that suggests code completions, functions, and entire files in your IDE.
Have an AI Tool?
List your AI tool for free, or go featured for top placement in your category - and reach thousands of potential users.
Submit Your Tool