szarkans/multi

23 stars · Last commit 2026-10-08

Use [Multi]ple AI agents for code-review, verifying and questions

README preview

<h1 align="center">multi</h1>

<p align="center"><a href="README.ru.md">[🇷🇺 →]</a> · <a href="README.zh.md">[🇨🇳 →]</a> </p>

<p align="center">run <code>multi</code>ple ai models for one task - code-review, planning, questions - and get <code>multi</code>ple opinions.</p>

---

<h2 align="center">what's this about?</h3>

`multi` runs one task through several AIs at once — best for code review ([measured: it finds more bugs than the built-in code review](#evals)).
it also does "is it actually done", [adhd](https://github.com/UditAkhourii/adhd) planning, or just a question — and shows you where the models converge and where they split. you judge, not them.

currently supports as `multi`-backends:

- claude subagents
- codex
- opencode
- kilo code (free models, no login)
- gemini

View full repository on GitHub →