RelayRouterDocs / Help / FAQ
FAQ
Common questions about RelayRouter.
#General
Do I need a new SDK?
No. The gateway speaks three protocols: point the OpenAI SDK at https://relayrouter.io/v1, point the Anthropic SDK at https://relayrouter.io (it calls /v1/messages), or use Gemini-style /v1beta endpoints. Swap your API key, done.
Which models do customers actually use?
The Claude line is the most popular, especially claude-opus-4-8 and claude-fable-5, alongside gpt-5.5. Gemini 3.5 is available too. The full live catalog with per-model rates is on the models page.
Am I charged for failed requests?
No. Only successfully completed calls are billed.
Is there a minimum spend or subscription?
No. Top up any amount and use it at your pace. No subscription required.
What payment methods are accepted?
Credit and debit cards via Stripe checkout.
How is pricing below official rates possible?
We negotiate volume pricing with providers and pass a portion of the savings to you. Mainstream model groups are priced 30% or more below official list prices; a small number of specialty groups can be at or above list. The live pricing page always shows the exact rate per model.
#API usage
How do I see which models my key can use?
Call GET /v1/models with your key, or browse the Models page. The per-key list returned by the API is the source of truth.
Are there rate limits?
Yes, standard rate limits apply per key. If you receive 429 Too Many Requests, slow down and retry with exponential back-off. Contact [email protected] if your workload needs higher limits.
How do I get support?
Email [email protected]. Include the request timestamp and, for async tasks, the task ID so we can trace it quickly.