Purpose: Create or edit 4–30 second short videos in the SdanceAI generator, suitable for scenarios such as product promotion, advertising creatives, app/SaaS explainers, creator brand kits, store promotions, and UGC-style scripts.
Key capabilities: Supports multimodal references from text prompts, first and last frames or up to 30 images, 10 videos, and 10 audio files; can generate synchronized dialogue, ambient sound, sound effects, and music; can reshape existing materials with reference videos; supports video extension, with a maximum output of 30 seconds; results can be exported as 480p or 720p MP4/MOV files.
How to use: In the generator, select the Seedance 2.5 model, enter prompts and reference materials, and set options such as aspect ratio, duration, resolution, and native audio. For API integration, create a key in the dashboard, send requests via Authorization: Bearer YOUR_API_KEY, call /api/ai-studio/execute to submit tasks (including modelId, visibility, and payload fields), then use /api/ai-studio/tasks/{taskId}?modelId={modelId} to poll task status (queued, running, succeeded, failed). The API is billed by credits, consistent with account balance and pricing rules.
Limitations: Current output is 4–30 seconds, 480p or 720p; first and last frame mode cannot be used simultaneously with multimodal reference arrays; uploaded files must comply with the format, size, duration, and aspect ratio limits shown in the generator; the API requires reasonable rate limiting, and generation stops when credits are insufficient; commercial use depends on platform terms and the rights to reference materials.
