Model catalog
Live models, pricing, and context windows — fetched from the gateway.
Each model has a public ID (e.g. gemini-2.5-flash) — a model offered at more than one context window is listed once per window, with the window appended to the ID (hypothetically, some-model-200k and some-model-1m for a model offered at both). Set that ID as model in a request.
The table below is live — it fetches GET /v1/models from the gateway. Prices are USD per 1M tokens and are the rates you're billed at.
Loading models…
Notes shown under a model id (e.g. "reasoning model — use max_tokens ≥ 1000") are the only capability hints we surface, because they're documented and verified. The /v1/models API does not expose vision/JSON/tool flags, so this catalog does not claim them. See Model notes.
Not everything callable is listed
This catalog is what we advertise, not the full set of IDs that work. Three kinds of model resolve and bill normally but are left out of it:
- Models whose output isn't text — image, video, audio and embedding models. This gateway is a text-completion service, and
/v1/chat/completionscannot return what they produce. - Models with no reported context window. Some upstreams don't publish one, and a window we can't state isn't something you can plan a request against.
- Router pseudo-models, which pick a model for you rather than being one. The
auto*entries are this gateway's own router and are listed.
If you already have one of these IDs, it keeps working. It simply won't appear in the table above or in GET /v1/models.