Seedance 2.5

What it is

Seedance 2.5 is an AI video generator built for creators.

Unlike tools where you write a prompt and hope for the best, Seedance 2.5 lets you create with your own material. Upload reference images to lock a character, a reference video to lock the camera move, audio to lock the rhythm — the model follows what you give it instead of improvising. Output runs up to 30 seconds in a single coherent take, with no watermark and a commercial-use license on every paid plan.

---

Core features

🎬 Text-to-video and image-to-video

Describe what you want and generate. Or supply a first frame and a last frame and let the shot travel from A to B. Prompts work in English and Chinese.

🧩 Omni-reference — where it really pulls ahead

Feed in all three kinds of material at once:

Reference type	Seedance 2.5	Seedance 2.0
Images	up to 30	up to 9
Videos	up to 10	up to 3
Audio	up to 10	up to 3

Reference a character's look, visual style, scene composition, camera movement, or transition timing. Seen a camera move you love? Drop the clip in and have it recreated with your own content. Seedance 2.5 can even generate from audio alone — beat-matched music videos in one shot.

🔊 Built-in sound and score

Flip the audio switch and the model generates matching voice, sound effects and background music from your prompt and the visuals — at no extra cost. Sound costs the same as silence.

🎛 Controls that stay out of the way

• Aspect ratio — 21:9, 16:9, 4:3, 1:1, 3:4, 9:16; landscape and vertical both covered
• Resolution — 480p, 720p, 1080p, plus 4K on Seedance 2.0
• Duration — up to 30 seconds on Seedance 2.5, 4–15s on the 2.0 series
• With first/last frames or a reference video, Seedance 2.5 matches the source aspect ratio automatically — no manual alignment

🚫 No watermark · Commercial use · Private

Finished videos carry no watermark. Every paid plan includes a commercial-use license. Your content is stored encrypted and is never made public or used for unrelated purposes.