Pricing
Pay for the tokens you use
Every model is billed per token at the provider's list price plus a routing fee. You prepay credits, and each request draws them down by what it used.
Cost-plus
Provider list price plus the same routing fee on every model, so we have no reason to steer you toward one.
One balance
A single credit balance covers every provider in the catalog, with a request cap on each API key.
Clean stops
When credits or a cap run out, the API returns 402 with a machine-readable code, not a retryable error.
Provider list prices
Read live from the catalog at https://api.gatewayz.ai/v1/models. Prices are per million tokens, before the routing fee.
| Model | Provider | Input | Output |
|---|---|---|---|
| Loading the catalog… | |||
How billing works
- Buy credits in the dashboard. Auto top-up is available if you'd rather not watch the balance.
- Each request is charged for the input and output tokens the provider reports. Prompt-cache reads are billed at the provider's cache rate where the provider offers one.
- An unknown model id returns
400 model_not_found. We never substitute a different model, so you're never billed for one you didn't ask for. - Out of credits returns
402 insufficient_credits; a spent per-key cap returns402 request_cap_exhausted.
Need invoicing or volume terms? Talk to us.