Frequently asked questions

Short answers to what people ask most often. If your question isn't here, message us in the site chat.

Do I need to rewrite my code?

No. The API is OpenAI-compatible: only the base URL and key change, everything else works as before — including the official Python and Node.js libraries.

Do I need any special network setup?

No, access goes through our base URL — no extra configuration required.

Is there a subscription fee?

No. Only actually used tokens are deducted, and there's no minimum monthly volume.

Does the balance expire?

No, the balance has no expiration date and is spent as you make requests.

How many keys can I create?

As many as you need: per project, per client, per environment. Each can have its own limit.

Can I switch models without reconnecting?

Yes, just change the model name in the request — the key and base URL stay the same.

Do the models search the internet?

Not on their own. Fresh data is added through your own search tool: results are passed back into the request.

What if my activation code isn't accepted?

Check for typos and extra spaces, then contact support via the site chat — we'll verify and activate the code.

For AI assistants

A machine-readable description of access and the model list lives at /llms.txt. Give this link to your assistant — it will fill in the base URL https://api.relaymodels.com/v1 and pick the right model for the task.

What's next