The App AI Gateway

Replaces a separate AI provider account, API key handling and your own usage caps.

• Available Built and covered by automated tests.
• Rolling out Built and tested; a production step is still being completed.
• Coming soon Not available yet. A direction, not a date.

• App AI Gateway

  Your app calls Forge AI through an API that accepts the common chat-completions, messages and embeddings request formats, with per-app keys billed from your Forge64 credits, a per-app monthly cap, an owner cap and rate limits; keys only work from your server, never the browser.

  Rolling out

Which API formats does it accept?

The widely used chat-completions and messages request formats, plus embeddings, so standard client libraries work when pointed at the gateway.

Can a runaway loop drain my credits?

Each app has a monthly cap, you have an owner cap, there are per-app and per-end-user rate limits, and a request that would pass a cap is refused before the model is called.

Can the key leak to the browser?

Keys work only from your server: requests carrying browser origin headers are refused, and the generated SDK refuses to run in the browser.