API docs rewritten
The API docs now match the live API: endpoints, request fields, async task polling and error codes, with separate image and video guides.
New models, API capabilities, Console improvements, and platform changes.
The API docs now match the live API: endpoints, request fields, async task polling and error codes, with separate image and video guides.
Reuse prompt in Assets opens the model page with the prompt, every setting, the route mode and the reference files already filled in. Review it and click Generate.
On the official route, image models are charged for the tokens each request actually used, at the official published rates, including text and thinking output. The official-cheap route charges half of those rates.
Seedance is billed per output token at the official list price on the official route and at 92% of it on the official-cheap route. Pricing pages show the rates with and without a reference video.
The website, playgrounds and Console are available in English, Simplified and Traditional Chinese, Japanese, Korean, German, French, Spanish, Brazilian Portuguese and Russian.
Drop reference images straight into a playground. Asset cards in the Console were redesigned to show the model, time and result at a glance.
Three more Nano Banana image models are available in the API and the playgrounds.
Seedance 2.5, 2.0, 2.0 Fast and 2.0 Mini are available, each in an international and a CN version, for text-to-video, image-to-video and reference video.
Seedream 4.0, 4.5, 5.0 and 5 Pro, plus GPT Image 2.5 Flare and Sunburst, joined the image catalog.
Set per-request and daily spending limits for your account. Send an Idempotency-Key when creating an async task so a retried request is not created or charged twice.
Balances and charges are shown in credits, at 1,000 credits per US dollar.
GPT Image 2, Nano Banana 2, Nano Banana Pro and Seedance generate through the API and the playgrounds, billed as you go.