# Letronna

> Letronna is an open-source AI agent, built in Rust, with its own cloud and wallet: it works on your computer and in your terminal with any model or account, hosts what it builds at a live link, and pays for the APIs it needs. macOS, Linux and Windows.

Letronna is made by Gadlus LLC. The agent runtime is Apache-2.0 on GitHub; hosted models and cloud deploy are the paid parts. Contact: hi@letronna.ai

## Key facts

- Runs locally on macOS 13+, Windows 11 and Linux (glibc 2.31+)
- Memory and context are plain-text files under ~/.letronna
- Works with any OpenAI-compatible endpoint, Anthropic, and local models via Ollama or llama.cpp
- No account needed for local use; nothing leaves the machine unless a hosted model is connected
- Capabilities: keeps session context, persistent memory on disk, local file and command access, scheduled runs, subagent delegation, MCP and shell tool integration

## What it does

- Desktop app (macOS, Windows, Linux) and a terminal app that share the same sessions and memory
- One app for many models: hosted models on a Letronna account, your own API keys (OpenAI, Anthropic, Gemini, xAI, OpenRouter), local models through Ollama, any OpenAI-compatible gateway of your own, and a ChatGPT plan's Codex models through a built-in plugin
- Routes across connected accounts by tier and falls back to the next one when an account hits its limit
- Letronna Cloud: runs a task on Letronna's servers from a prompt, a repository or a folder on your computer, and keeps web apps and bots (Discord, Telegram, schedulers) online around the clock
- Tools for files, shell commands, web search and fetch, a browser, subagents ("crew"), schedules and MCP servers; less used tools are loaded on demand to keep each request small
- Written in Rust as one native binary

## Benchmarks

- SWE-bench Verified, the 10 tasks of OpenHuman's harness benchmark, graded by the official evaluator, on DeepSeek V4.1 Flash: Letronna solved 10 of 10 at 376k tokens and $0.0081 per solved task, the fewest tokens of any harness that solved all ten (OpenCode 390k, Claude Code 428k, Hermes 482k). Details and per-task data: https://letronna.ai/benchmark
- Harness-Bench (Qihoo360, 106 agent tasks, https://arxiv.org/abs/2605.27922), rerun 13 October 2026: Letronna scored 84.0 on DeepSeek V4 Flash, 83.6 on Muse Spark 1.3, 83.4 on Claude Opus 5.5 and Claude Sonnet 5.5, and 80.4 on Claude Sonnet 4.6. The paper's best configurations, all on GPT-5.4: NanoBot 81.3, Codex 80.4, Hermes 74.0. On the same model, Letronna beat the paper's average harness by 15.1 points on Claude Sonnet 4.6 and 26.5 on DeepSeek V4 Flash, and used 47K tokens and 6.4 model calls a task on Claude Sonnet 5.5 (paper means: Codex 86K and 5.0, Hermes 140K and 22.6)

## Pricing

- Free: $0 USD forever - Bring your own keys.
- Pro: $20 USD per month - $20 of hosted credit a month, or $240 at once on the yearly plan, and lower prices on every hosted model.
- Team: $18 USD per seat per month, 2 to 5 seats - One credit pool for the whole team.

## FAQ

**Does it need an account?**
No. The app runs locally with your own model keys. An account only exists for hosted models and cloud deploy on Pro.

**Where does my data live?**
On your disk. Context and memory are plain-text files under ~/.letronna. Nothing leaves the machine unless you connect a hosted model.

**Which models can I use?**
Any OpenAI-compatible endpoint, Anthropic, local models through Ollama or llama.cpp, and the hosted pool on Pro.

**Does it keep memory between sessions?**
Yes. Memory is written to disk as you work and loaded on the next launch. Edit or delete it like any other file.

**What happens when I quit?**
Scheduled runs pause unless you have deployed to cloud. Nothing runs on your machine while the app is closed.

**Is it open source?**
The agent runtime is Apache-2.0 on GitHub. The hosted model pool and cloud deploy are the paid parts.

## Links

- [Benchmark](https://letronna.ai/benchmark)
- [Docs: getting started](https://letronna.ai/docs/getting-started)
- [Docs: the app](https://letronna.ai/docs/app)
- [Docs: the terminal](https://letronna.ai/docs/terminal)
- [Docs: gateways and models](https://letronna.ai/docs/gateways)
- [Docs: hosted models](https://letronna.ai/docs/hosted-models)
- [Docs: Letronna Cloud](https://letronna.ai/docs/cloud)
- [Docs: teams](https://letronna.ai/docs/teams)
- [Docs: data and privacy](https://letronna.ai/docs/data-and-privacy)
- [Changelog](https://letronna.ai/changelog)
- [Roadmap](https://letronna.ai/roadmap)
- [Partner with Letronna](https://letronna.ai/partner)
- [Sitemap](https://letronna.ai/sitemap.xml)

- [Home](https://letronna.ai/)
- [Capabilities](https://letronna.ai/#capabilities)
- [Platforms](https://letronna.ai/#platforms)
- [Pricing](https://letronna.ai/#pricing)
- [FAQ](https://letronna.ai/#faq)