FAQ
Answers to common questions.
How much does it cost?
The per-token price listed on each model, paid from prepaid credits, plus a 5.50% fee when you buy credits. See Pricing.
Is there a free tier?
No. Every request is paid from credits. Purchases start at $5.00.
Do you store my prompts and completions?
Our request log stores metadata (model, tokens, cost, latency, status), not prompt or completion text. Content is sent to the upstream provider that serves the model, which processes it under its own terms. See the Privacy Policy.
Which provider serves my request?
We route each request to an endpoint we have authorized for that model and fail over to another eligible endpoint if one fails. The price you pay is the model's listed price regardless of which endpoint serves it.
Why did I get 402 when I have credits?
Requests hold their maximum possible cost first. Without max_tokens, a non-streaming request holds the model's full output limit; a streaming request holds a first slice and grows it as output arrives. Stream, lower max_tokens, or add credits. See Billing.
Can I bring my own provider keys?
Not today. All requests run on our provider accounts and are billed from your credits.
How do I get support?
Email support@aimodelsrouter.ai with your generation id (gen-…) if your question is about a specific request.