Seedance 2 API (poyo.ai)

Seedance 2 API generates a 4–15 second video from text, up to two frames, or image, video, and audio references. It offers a faster variant limited to 480p and 720p, and is available on PoYo. The model supports text-to-video, image-to-video, multimodal reference-to-video, and audio-driven video workflows, with native audio generation and control over first/last frames.

Seedance 2 API serves content creators and developers who need short-form, cinematic video generation with reference integration. It supports multi-shot storytelling, director-level control, and native audio, aiming at workflows such as pre-visualization, localization, advertising, gaming, education, and music visuals.

The API emphasizes an async approach with webhook callbacks, and a unified model capable of handling multiple input modalities in a single request. It provides endpoints for submitting tasks, following status, and retrieving results, all within PoYo’s tooling ecosystem.

Seedance 2 API offers standard and fast variants, with input schemas shared across modes and constraints on resolution, duration, and file references. The notes describe model options, asset limits, and how to use image, video, and audio references in different generation modes.

Use cases

• Film and Pre-Visualization: Generate multi-shot pre-vis sequences with synchronized dialogue and ambient audio from a single prompt.
• Multilingual Content Localization: Localize with native lip-sync across languages and replace manual dubbing.
• Advertising and Marketing: Rapidly iterate cinematic ad concepts using references and audio.