Seedance 2 Fast is the production-oriented option for teams that want quicker and lower-cost iteration across text-to-video and image-to-video. The MakeInfluencer contract supports 4–15 second clips, 480p through 1080p, generated audio, an optional end frame, and reference images.
Model Contract
| Field | Behavior |
|---|---|
prompt | Required scene and motion direction |
startFrameUrl | Optional; switches to image-to-video |
endFrameUrl | Optional last-frame guidance |
aspectRatio | Defaults to 16:9 for text generation |
duration | 4–15 seconds, default 5 |
resolution | 480p, 720p, or 1080p |
generateAudio | Defaults to true |
reference_images | Optional reference image array |
The Seedance 2 Fast API reference reflects the live registry and exact credit quote.
Example Request
curl --request POST \
https://www.makeinfluencer.ai/api/v1/models/seedance2fast/preview-credits \
--header "Authorization: Bearer $MAKEINFLUENCER_API_KEY" \
--header "Content-Type: application/json" \
--data '{
"input": {
"prompt": "A creator turns toward camera on a busy night street, slow push-in, natural traffic ambience",
"startFrameUrl": "https://example.com/approved-creator.jpg",
"duration": 5,
"resolution": "720p",
"generateAudio": true
}
}'
The preview accounts for duration, resolution, and whether references are present. That is important because reference-assisted generation can use a different provider rate.
When to Use 480p, 720p, or 1080p
480p for Iteration
Use 480p to test motion prompts, camera behavior, and whether the source image animates cleanly. A draft that fails creatively does not become better because it was rendered at 1080p.
720p for Working Content
Use 720p for internal review, fast social experiments, and workflows where output speed and cost matter.
1080p for Approved Directions
Use 1080p after the prompt, source image, and motion have already succeeded. Preview the exact credit difference before promoting a batch.
Escalate Resolution, Not Uncertainty
Find the winning motion at the lowest acceptable resolution. Re-render only approved directions at the delivery quality your customer needs.
Start and End Frame Workflow
An end frame can guide where a shot should arrive, but it also creates a harder constraint. Use it when the final composition matters, such as:
- A product reveal that ends on a pack shot
- A character moving from a wide frame to a close-up
- A transition between two approved campaign poses
- A loop designed to return toward a known visual state
Keep the start and end images compatible in identity, clothing, lighting, and scene geometry. Large contradictions force the model to invent an implausible transition.
Generated Audio vs Controlled Speech
Set generateAudio when ambient sound or model-generated audio benefits the clip. For exact spoken content, use a separate audio-driven workflow such as the InfiniteTalk API, where the uploaded audio controls timing and speech.
Seedance 2 Fast vs Seedance 2.5
Seedance 2 Fast is live and available through the current registry. Seedance 2.5 has been unveiled, but production API access is still being verified. Track the Seedance 2.5 API availability page for the integration status.
Do not block a product launch on an unverified future endpoint. Keep the model ID configurable, ship on a production model, and compare 2.5 when its pricing and reliability are real.
API FAQ
Does Seedance 2 Fast support text-to-video?
Yes. Omit startFrameUrl and send a prompt plus the desired aspect ratio, duration, and resolution.
Does it support image-to-video?
Yes. Add startFrameUrl; you can also provide an end frame where supported by the model contract.
Does reference usage affect price?
It can. The registry has separate per-second pricing tables with and without references. Always call preview-credits with the exact planned input.
Is Seedance 2.5 already available?
Not through MakeInfluencer as of August 5, 2026. The planned integration will happen after production provider access and verification.