Native 30-Second Generation
Create longer clips in one pass for demos, short ads, and narrative beats without manual stitching.
Powered by ByteDance Seedance 2.5
Seedance 2.5 Video Generator – 30-Second Native Audio, Large Reference Sets
Generate up to 30-second videos with Seedance 2.5. Use up to 30 reference images, 10 reference videos, and 10 reference audios with native synced audio in one workflow.
No creation yet
Submit the form to start a new creation.
Seedance Version Comparison
Compare Seedance 2.5, the 2.0 family, and Seedance 1.5 Pro, 1 Pro, 1 Pro Fast, and 1 Lite.
| Model | Max duration | Resolution | Reference images | Reference videos | Reference audios | Best for |
|---|---|---|---|---|---|---|
| Seedance 2.5 Flagship production model | 30s (Auto supported) | 480p, 720p | 30 | 10 | 10 | Long-form clips, complex reference sets, production delivery |
| Seedance 2.0 High-quality multimodal video | 15s (Auto supported) | 480p, 720p, 1080p | 9 | 3 | 3 | 1080p final renders and reference-heavy creative control |
| Seedance 2.0 Fast Faster iteration variant | 15s (Auto supported) | 480p, 720p | 9 | 3 | 3 | Prompt testing, batch exploration, quick drafts |
| Seedance 2.0 Mini Lower-cost high-volume variant | 15s | 480p, 720p | 9 | 3 | 3 | High-volume generation with the same multimodal workflow |
| Seedance 1.5 Pro Cinematic audio-video model | 12s | 480p, 720p, 1080p | First and last frame | — | Optional synced audio | Cinematic text-to-video and image-to-video with optional generated audio |
| Seedance 1 Pro Pro text and image to video | 12s | 480p, 720p, 1080p | First and last frame | — | — | 1080p cinematic clips with a controllable camera |
| Seedance 1 Pro Fast Faster Seedance 1 Pro | 12s | 480p, 720p, 1080p | First frame | — | — | Quicker, lower-cost drafts from a prompt or one image |
| Seedance 1 Lite Lightweight reference model | 12s | 480p, 720p, 1080p | 4, or first and last frame | — | — | Lower-cost clips guided by up to four reference images |
What this workflow is good at
Large Reference Sets
Combine up to 30 images, 10 videos, and 10 audios for character consistency, motion transfer, and audio-led edits.
Native Synced Audio
Generate dialogue, ambience, and effects together with the visuals for lip-sync and beat-aware scenes.
Long-form product demos
Use a single 20–30 second render for feature walkthroughs instead of chaining shorter clips.
Reference-heavy ad variants
Keep character, product, motion, and audio cues stable across multiple campaign outputs.
Presenter and dialogue clips
Generate spoken lines with native audio and stronger instruction following for camera and performance.
How it works
1
Step 1: Write a production-ready prompt
Describe subject, motion, camera behavior, and audio intent. Put dialogue in double quotes and call out references like [Image1] or [Video1] when you upload assets.
2
Step 2: Add the references you need
Upload reference images for identity and style, reference videos for motion or editing intent, and reference audios for lip-sync or rhythm. Keep first/last frame mode separate from reference-image mode.
3
Step 3: Choose duration and render
Pick 5 to 30 seconds or Auto, choose 480p or 720p, then generate with native synced audio enabled.
How to use it
Seedance 2.5 is ByteDance's flagship multimodal video model for creators who need longer clips, richer reference control, and tighter alignment between dialogue, motion, and sound in a single generation.
The headline upgrade is native 30-second generation. Instead of stitching shorter outputs together, Seedance 2.5 can produce a full half-minute clip in one pass. That makes it a better fit for product demos, presenter clips, short ad variants, and reference-heavy creative direction.
Reference capacity also expands significantly. Seedance 2.5 supports up to 30 reference images, 10 reference videos, and 10 reference audios in the same request. That is useful when you need character consistency, motion transfer, edit intent, and audio timing cues at the same time.
Audio is still generated natively with the visuals. Dialogue, ambience, effects, and music can be created together, which keeps lip-sync and pacing aligned from the first render. For spoken lines, double quotes in the prompt remain the most reliable cue.
Seedance 2.5 does not replace every 2.x use case. Seedance 2.0 remains the better choice when you need 1080p output. Seedance 2.0 Fast and 2.0 Mini remain useful when speed or cost matters more than maximum reference capacity or clip length.
Keep exploring Seedance
- Seedance 2.5 Generator
Generate up to 30-second clips with larger reference sets and native synced audio.
- Seedance 2.0 Family
Generate with Seedance 2.0, 2.0 Fast, and 2.0 Mini in one shared studio workflow.
- AI Video Generator
Switch between Seedance 2.5 and the Seedance 2.0 family in the studio.
- Seedance 2.5 vs 2.0
Read a practical comparison of duration limits, reference caps, and output quality.
- Seedance 2.0 vs 1.5 Pro
See how the multimodal 2.x family differs from the earlier cinematic baseline.
- Seedance 2.0 Prompt Examples
Review prompt patterns for dialogue, references, and audio-led scenes.
Frequently Asked Questions
Start generating
Generate up to 30-second videos with Seedance 2.5. Use up to 30 reference images, 10 reference videos, and 10 reference audios with native synced audio in one workflow.