Thirty seconds in one take
A single generation runs the full thirty seconds, so identity, wardrobe and props stay fixed from frame 1 to frame 900 instead of drifting across stitched clips.
ByteDance
Seedance 2.5
Native 30-second video, 50 multimodal references. A single 30-second take that holds identity, clothing and props from the first frame to the last, steered by as many as fifty image, video and audio references.
This exact endpoint is not in the Studio catalog yet — the buttons open our text-to-video and image-to-video workflows, where comparable models are available today.
- Max duration
- 30s
- Max resolution
- 720p
- Audio
- Native
- Aspect ratios
- 21:9, 16:9, 4:3, 1:1, 3:4, 9:16, 9:21
- References
- Up to 50 multimodal
- Commercial use
- Allowed
Features
What Seedance 2.5 does differently.
Fifty multimodal references
Up to thirty images, ten videos of thirty seconds each and ten audio tracks of thirty seconds each can be supplied in one pass to lock characters, motion and sound.
3D layout references
Block a scene in a 3D tool, export the viewport and hand it over as layout. Camera position and staging follow the blockout rather than a written description.
Green screen editing
Drop a subject into a new plate and it reacts to the replacement lighting and physics instead of sitting flat against the background.
Timestamp pacing and eleven languages
Beats can be pinned to timestamps so a thirty-second cut lands where you want it, and dialogue is native in eleven or more languages including Chinese, English, Spanish, Indonesian, Malay, Thai, Arabic, Portuguese, Vietnamese, Japanese and Korean.
Examples
Prompts that show its range.
Circus long take: Victorian circus magician performers, choreographed chaos, continuous
Try this promptSports short film: Michael Owen career retrospective, signature moves
Try this promptCorporate comedy: boardroom, lost sock hunt and washing machine POV dive, dialogue and subtitles
Try this promptAnimated short: otter courier, stag beetle bus, tapir bubble plane, 60+ shots across 4 lighting environments
Try this promptThese clips were produced with Dreamina Seedance 2.5 and published by BytePlus. Run a prompt in the Studio to see your own render.
Workflows
Every way to run Seedance 2.5.
- Text to videoPrompt to video, audio in the same passOpen workflow
- Image to videoAnimate a still into a continuous takeOpen workflow
- Reference to videoImage, video and audio references, up to fifty inputsOpen workflow
ByteDance has not published a public rate for Seedance 2.5, so treat any figure quoted elsewhere as unconfirmed. Credits come from one shared balance across image, video and speech, and the exact estimate is shown in the Studio before you generate.
Built for
- Full thirty-second ad spots cut once per aspect ratio
- 3D blockout to finished render with timestamp-controlled pacing
- On-brand creator content in one pass across eleven or more languages
- Green screen replacement where the subject has to sit in new lighting
FAQ
Seedance 2.5, answered.
Limits, audio, licensing and price — the things worth knowing before you spend a credit.
Thirty seconds, one take.
Open the Studio to run a comparable model today, and we will wire Seedance 2.5 in as soon as the endpoint lands in our catalog.