Cursor

Cursor lets you bring your own OpenAI key for its chat and agent, and send those requests to a different base URL. Point it at Inferray to use any model in the catalogue on your Inferray balance.

Set it up

1. Sign in to Cursor and open Cursor Settings → Models → API Keys.
2. Under OpenAI API Key, paste your Inferray key.
3. Turn on Override OpenAI Base URL and enter https://api.inferray.com/v1.
4. Click Verify.
5. Under Models, add a custom model with an ID from the catalogue, such as claude-sonnet-4-6 or gpt-5.5, and turn it on.

Select that model explicitly in a new chat and send a short prompt. The request appears in your logs within a moment.

What uses your key

Your key is used for chat and agent requests with the custom model. Tab completion and some background features keep using Cursor's own models, and Cursor still needs its own account sign-in. Requests made with your key are billed to your Inferray project, not your Cursor plan.

Troubleshooting

• Verification fails — check the base URL ends in /v1 and the key is active.
• A model is rejected — copy its ID exactly from the catalogue; use a model with tool support for agent mode.
• Other providers stop working — the OpenAI base URL override applies to every OpenAI-format model in Cursor. Turn it off to go back.