Seedance 2.5
ByteDance Seedance 2.5 - Up to 30s clips from text, images (with end-frame control) & references, audio included. On Influverse a 5s clip at 480p costs 42 credits, charged from the same wallet as every other model. There is no separate subscription for this model and no training step to use your own character with it.
By Influverse AI Team · Last verified: 6 September 2026 · Editorial policy
Generate with Seedance 2.5Capabilities
| Vendor | ByteDance |
|---|---|
| Duration | 4 to 30 seconds |
| Resolutions | 480p, 720p |
| Audio | Native audio on every generation |
| Text to video | Yes |
| Image to video | Yes |
| First and last frame | Yes |
| Motion control | No |
| Extend an existing clip | No |
| Reference video | Yes |
| Reference images | Up to 9 |
What it costs in credits
Derived from the same pricing source the app charges against, so these figures and the cost preview on the generation screen can never disagree.
| Duration | Resolution | Credits |
|---|---|---|
| 4s | 480p | 34 |
| 5s | 480p | 42 |
| 10s | 480p | 84 |
| 30s | 480p | 252 |
| 4s | 720p | 72 |
| 5s | 720p | 90 |
| 10s | 720p | 180 |
| 30s | 720p | 540 |
Credits come from your plan and from credit packs. See the pricing page for the plan ladder.
When to use Seedance 2.5
Seedance 2.5 is the model to reach for when the shot needs to be longer than a normal social clip and still hold together. It is the only model in the catalogue that will take a single generation out to thirty seconds, and it accepts a start frame, an end frame, a stack of reference images and reference video at the same time, so a single call can carry a character, a product and a movement style into one continuous take. Audio is generated with the clip rather than bolted on afterwards, which matters when the motion and the sound have to agree, for example footsteps on gravel or a hand setting a cup down. Pick it when you are building a hero clip that has to carry a story beat on its own: an opening scene for a reel, a product demo where the camera moves around the object, or a walk-and-talk that would need three separate generations on a shorter model. Pick something else when you only need four or five seconds of B roll, because you are paying for a long-form engine to do a short job and the per second cost is the same either way. The two resolution tiers behave very differently on price, so draft at the lower tier until the prompt is right and only then rerun at the higher one. It is also the model we would choose when a scene has to start on one image and finish on another, since end frame control is native rather than approximated with a second generation.
Strengths
- Longest single take in the catalogue
- Start frame and end frame control in one generation
- Accepts reference images and reference video together
- Audio generated with the clip, not layered afterwards
Limits
- Only two resolution tiers, and the higher one costs materially more per second
- Long generations take longer to come back, so iteration is slower
- No motion transfer from a reference clip, use Kling Motion Control for that
Best for
- Hero opening shots for a reel
- Product demos with camera movement
- Scenes that must begin and end on specific frames
Other models to consider
Back to the full model catalogue
More from Influverse
Generate with Seedance 2.5
Every new account starts with a free trial that unlocks every surface and every model. No card required.
Start creating free