FAQ

Is it really free?

Yes — forever free to start. The full IDE and an on-device model are free with an allowance that refills every 5 hours. Pro ($3.99/mo) removes limits; it doesn’t unlock the product.

Is my code uploaded somewhere?

No. Generation runs on your device; your files stay in your browser or your local folder. The exact accounting — including the few things that do touch a server (login, token counts, billing) — is in Privacy by design.

Do I need a GPU?

No, but it helps. With a GPU, OIOXO runs a bigger on-device tier via WebGPU. Without one, the Basic tier runs on CPU (slower), or you can plug in your own API key and skip local inference entirely.

Can I use GPT/Claude/Gemini instead of the local model?

Yes — 12 providers plus any OpenAI-compatible endpoint and local Ollama. Your key stays in your browser, and OIOXO’s prompt-caching typically cuts your provider bill substantially. See Bring your own key.

What can it build?

Web pages, games, React apps, APIs, CLI tools, Python scripts, SQL — scaffolded, run, and previewed in the browser. For anything with a visible result you get a live preview and click-to-edit.

Does it work offline?

Once your model is downloaded and cached, code generation itself runs without the network. Account features, GitHub, publishing, and BYOK naturally need a connection.

How is this different from Cursor or Copilot?

Two ways. First, OIOXO verifies: it runs your code — compiler and live preview — and repairs failures before calling anything done. Second, it’s private by architecture: the default model runs on your device, so there is no server reading your code. And it’s a full IDE in a tab — nothing to install.

Can I work on my existing projects?

Yes — open a local folder on the Files surface (Chromium), or open a repository on the GitHub surface and push changes back.

Who owns what I build?

You do. Code you create with OIOXO is yours — export it, publish it, sell it. See the Terms of service.