# Letta review

> Letta is a platform for building stateful AI agents with memory that can learn and improve over time.

Devtoolwatch Board, the week of Oct 25–31, 2026. Canonical page: https://index.devtoolwatch.com/tools/letta/

## TL;DR

Letta, formerly MemGPT, is a platform for building stateful agents whose memory persists and improves with use. Letta is for developers who want coding agents, personal assistants, AI coworkers in chat apps, or stateful agents embedded in TypeScript applications. Letta's agent harness is open source and runs in a terminal UI, a self-hosted App Server, the web, or Letta Cloud.

## Letta at a glance

- **Category:** [RAG, data and memory](https://index.devtoolwatch.com/categories/rag-data-and-memory/)
- **Maintained by:** Letta (organization, [@letta-ai](https://github.com/letta-ai) on GitHub)
- **License:** Apache 2.0
- **Repository:** [github.com/letta-ai/letta](https://github.com/letta-ai/letta)
- **Website:** [docs.letta.com](https://docs.letta.com/)
- **npm package:** [@letta-ai/letta-client](https://www.npmjs.com/package/@letta-ai/letta-client) 1.12.1
- **PyPI package:** [letta](https://pypi.org/project/letta/) 0.34.12
- **Latest release:** [letta 0.34.12](https://pypi.org/project/letta/) on PyPI, Oct 12, 2026
- **First published:** Oct 11, 2023

## What Letta does

- Letta builds stateful agents with memory that persists and can learn and improve over time.
- Letta provides an interactive terminal UI, launched with the letta command, for working with agents locally.
- Letta runs an App Server for local or self-hosted agents.
- Letta connects agents to Slack, Telegram, Discord, and custom channels.
- Letta offers an Agent SDK for building agents into TypeScript applications.
- Letta supports Skills for reusable workflows, Schedules for recurring prompts, and Mods for extending the harness.
- Letta Cloud keeps agent memory, identity, and conversations available across computers.

## How to install Letta

**With npm**

```
npm install -g @letta-ai/letta-code
```

### Requirements

- npm, to install the @letta-ai/letta-code package.
- An LLM provider configured for your agents.
- Python 3.9 or later for the letta PyPI package.

### First steps

1. Install Letta globally from npm with the @letta-ai/letta-code package.
2. Run letta to launch the interactive terminal UI.
3. Choose an LLM provider, then follow the quickstart to chat with the Tutor agent.
4. Run letta server to host agents locally or on your own infrastructure.

## Who Letta is for

- Developers embedding stateful agents with persistent memory into TypeScript applications.
- Developers who want a memory-first coding agent in the terminal.
- Teams running AI coworkers in Slack, Telegram, or Discord.
- Users automating recurring work with schedules and reusable skills.

### Limitations

- Letta's current source lives in letta-ai/letta-code; the letta-ai/letta repo's V1 API server is retired and kept on an archive branch.
- Letta channels for local-backend agents require keeping the channel server running on infrastructure you manage.
- Letta Server 0.16.7 removed block limit enforcement as a breaking change.

## Usage data (the week of Oct 25–31, 2026)

- Board rank: #362 overall, #36 in RAG, data and memory, momentum score 17
- Downloads that week: 116k (−16% vs 4-week average)
- New GitHub stars per week (last 12 weeks): Aug 9: 115; Aug 16: 97; Aug 23: 144; Aug 30: 142; Sep 6: 95; Sep 13: 91; Sep 20: 99; Sep 27: 133; Oct 4: 95; Oct 11: 95; Oct 18: 89; Oct 25: 77

## Letta alternatives

| Rank | Tool | Stars | Downloads/wk |
|---|---|---|---|
| #1 | [XERJ](https://index.devtoolwatch.com/tools/xerj-org-xerj/) | 17.9k | – |
| #2 | [Memvid](https://index.devtoolwatch.com/tools/memvid-memvid/) | 17.3k | – |
| #3 | [LEANN](https://index.devtoolwatch.com/tools/startrail-org-leann/) | 13.7k | – |
| #4 | [Graphify](https://index.devtoolwatch.com/tools/graphify-labs-graphify/) | 140k | 480.7k |
| #5 | [MarkItDown](https://index.devtoolwatch.com/tools/markitdown/) | 193.2k | 4.7M |
| #6 | [Firecrawl](https://index.devtoolwatch.com/tools/firecrawl/) | 197.4k | 2.9M |

## Letta FAQ

### Is Letta free and open source?

Yes. Letta's source code is on GitHub (letta-ai/letta) under the Apache 2.0 license, so you can use it for free. Its website also links to a pricing page (https://docs.letta.com/pricing) for any paid plans.

### How do I install Letta?

Run `npm install -g @letta-ai/letta-code`. The install section above has the other options from the project's README.

### Is Letta still maintained?

Yes. The letta-ai/letta repository was last updated on Oct 18, 2026, and the latest release, letta 0.34.12 on PyPI, came out on Oct 12, 2026.

### How popular is Letta?

Letta has 25.3k GitHub stars as of Nov 1, 2026. It gained 77 new stars in the week of Oct 25–31, 2026, against a 12-week average of 106 a week. Its main package had about 116k downloads that week, −16% against its own 4-week average. That made it #36 among RAG, data and memory tools on the Devtoolwatch Board.

### Letta vs Mem0: how do they compare?

Letta and Mem0 are both tracked by Devtoolwatch as RAG, data and memory tools. Letta has 25.3k GitHub stars and Mem0 has 68.2k. In the week of Oct 25–31, 2026, Letta's main package had 116k downloads and Mem0's had 884.1k. Mem0 is a memory layer for AI agents and apps that stores and retrieves user, session, and agent context across interactions. [Mem0 review](https://index.devtoolwatch.com/tools/mem0/)

### Letta vs LangGraph: how do they compare?

Letta and LangGraph are both tracked by Devtoolwatch. Letta has 25.3k GitHub stars and LangGraph has 44.2k. In the week of Oct 25–31, 2026, Letta's main package had 116k downloads and LangGraph's had 18.5M. LangGraph is a low-level orchestration framework and runtime for building, managing, and deploying long-running, stateful agents in Python or JavaScript. [LangGraph review](https://index.devtoolwatch.com/tools/langgraph/)

### What are the best alternatives to Letta?

Other RAG, data and memory tools on the Devtoolwatch Board in the week of Oct 25–31, 2026, by momentum: XERJ (#1), Memvid (#2), LEANN (#3), Graphify (#4), MarkItDown (#5). The alternatives table above has their numbers.

### Is Letta the same as MemGPT?

Yes. Letta was formerly known as MemGPT. The project is actively developed, and its current source code lives in the letta-ai/letta-code repository.

### Can I self-host Letta?

Yes. Letta's App Server runs local or self-hosted agents with the letta server command. Letta Cloud is the managed alternative for keeping memory and conversations available across computers.

### Is Letta open source?

Yes. The Letta agent harness is fully open source, according to the Letta documentation.

### What can I build with Letta?

Letta builds coding agents, personal assistants, AI coworkers in chat apps, and stateful agents embedded in your own applications.

### What happened to the Letta V1 API server?

The Letta V1 API server is retired. It remains on the archive branch with existing tags and releases for reproducibility, but active projects should use letta-code.

## Sources

1. [letta-ai/letta on GitHub](https://github.com/letta-ai/letta)
2. [Letta README](https://github.com/letta-ai/letta/blob/main/README.md)
3. [@letta-ai/letta-client on npm](https://www.npmjs.com/package/@letta-ai/letta-client)
4. [letta on PyPI](https://pypi.org/project/letta/)
5. [Letta website](https://docs.letta.com/)
6. [Release 0.16.8 (May 14, 2026)](https://github.com/letta-ai/letta/releases/tag/0.16.8)
7. [Release 0.16.7 (Mar 31, 2026)](https://github.com/letta-ai/letta/releases/tag/0.16.7)
8. [Release 0.16.6 (Mar 4, 2026)](https://github.com/letta-ai/letta/releases/tag/0.16.6)
9. [Hacker News: Letta: Letta is a framework for creating LLM services with memory](https://news.ycombinator.com/item?id=43294974)
10. [Hacker News: Letta Code: a memory-first coding agent](https://news.ycombinator.com/item?id=46293374)
11. [Devtoolwatch Board, the week of Oct 25–31, 2026](https://index.devtoolwatch.com/)