CodeMouse

Yes — CodeMouse runs on your own Anthropic, OpenAI, and/or Gemini key (any one is enough). You pay your provider directly for inference, about $0.05–0.15 per review — we never mark up tokens. You add your key in the dashboard after installing; it's encrypted at rest and never shared.

Yes. One flat lifetime price — no per-seat fees, no usage caps, and unlimited repositories, pull requests, and team members. The only thing you ever pay beyond the deal is your own model provider for inference, on your own key.

Two things. It runs multiple models (Claude, GPT, Gemini) in parallel and combines their findings, so a blind spot in one is caught by another. And it reads the existing comments on a PR before it speaks, so it never repeats your teammates, other bots, or its own past runs — no echo, no noise.

Claude (Anthropic), GPT (OpenAI), and Google Gemini — one, two, or all three running in parallel, your choice. CodeMouse installs as a GitHub App and reviews pull requests on any repository.

One-click: install the GitHub App, pick your repositories, add your API key, and open a pull request — CodeMouse reviews it automatically within minutes. It works with any language and any GitHub repo, public or private. No CI pipelines or YAML to configure.

Your repository is cloned only for the review and deleted immediately afterward — your code is never stored. Your API keys are encrypted at rest (AES-256-GCM) and used only to run your reviews; they're never shared.