Seedance 2.5 AI Video Generator

Generates up to 30 seconds of video in a single generation, doubling the previous 15-second limit for more coherent motion and narrative flow.

Supports up to 50 reference assets including images, videos, and audio in one generation to orchestrate multiple elements into a single scene.

Produces videos at up to 3840 × 2160 resolution natively without external upscaling, suitable for direct use in professional content pipelines.

Approximately 20% better adherence to subjects, actions, layout, and style instructions compared to Seedance 2.0, reducing regeneration needs.

Accepts text, image, video, and audio inputs simultaneously to generate synchronized video and audio output in one pass.

Generates native audio tracks synchronized with the video output for complete multimedia content.

Supports various aspect ratios including 16:9, 9:16, 1:1, 3:4, and others to match different delivery channels.

Accepts text prompts in both Chinese and English for diverse user needs.