Every AI coding tool I used had the same two problems: a monthly bill that never stops, and a usage limit or a refusal right when I needed it most. So I built Grunz — a desktop AI agent you own outright.
What it is
Grunz is a native app for Mac, Windows, and Linux that behaves like an agentic coding assistant. You give it a goal and it plans, writes and edits files across a whole repo, runs commands, watches the output, fixes itself, and hands you the finished thing. It also does cited web research and generates images. One app instead of five tabs.
The part that's different
You bring your own model endpoint. Grunz doesn't host a model — you connect one (the setup guide spins up a pay-as-you-go endpoint on Featherless), and from then on your prompts, files, and conversations stay on your machine. Because it runs open models, there are no weekly caps and no "I can't help with that."
Under the hood
- A tool loop with ~50 tools: read/write/edit files, run shell commands, start dev servers, a headless-Chrome browser, image generation, MCP server support, and subagents.
- A plan mode that proposes before it touches anything.
- A permissions dial: ask before every change, auto-accept edits, or full auto.
What it costs
$100 once. No subscription. You pay your model provider directly for compute, and that's the only ongoing cost — which means you control it.
Honest status
Early version. The Mac build isn't Apple-notarized yet, so the first launch needs a right-click > Open. Desktop is what ships today.
If you've been looking for a Claude-Code-style agent that runs on open models and doesn't rent you your own tools, take a look: https://grunzai.com
I'd genuinely like feedback on the agent loop and the model handoff.
Top comments (0)