Six models. One controlled API.
Only these exact model versions are available. Unsupported models and parameter combinations are rejected before a provider request.
| Model | Type | Key strengths | Input | Output |
|---|---|---|---|---|
Seedance 2.0dreamina-seedance-2-0-260128 | Video | High quality, general purpose | Text, image, audio, video | Video |
Seedance 2.0 Fastdreamina-seedance-2-0-fast-260128 | Video | Faster generation, lower cost | Text, image, audio, video | Video |
Seedance 2.0 Minidreamina-seedance-2-0-mini-260615 | Video | Lightweight, cost efficient | Text, image, audio, video | Video |
Seedance 2.5dreamina-seedance-2-5-260628 | Video | Higher fidelity, complex motion | Text, image, audio, video | Video |
Seedream 5 Prodola-seedream-5-0-pro-260628 | Image | Photoreal and style control | Text, up to 10 images | Image |
Seed Audio 1.0seed-audio-1.0 | Audio | Natural speech and sound | Text, audio OR image | Audio |
Capabilities, without ambiguity.
- Seedance 2.5 Draft → Final
- Draft is 480p. Converting a Draft produces 1080p Final, with the original input-video pricing facts inherited. Draft and Final are two separately billed generations.
- Seedream 5 Pro
- Single-image generation, up to ten reference images and interactive editing. Layer decomposition, sequential generation and streaming are not enabled.
- Seed Audio 1.0
- Audio references or one image reference, sound configuration and subtitles. Audio and image references cannot be mixed. Actual original duration determines the bill.
We serve the uncensored creative market. Provider content policies still apply; we do not promise moderation-free generation or bypass provider restrictions.