The workflow combines a portrait with a fixed motion reference. Seedance Mini creates a new video guided by those inputs; results vary, so a final review matters as much as your setup.
1. Choose a portrait that gives the model a clear subject
Start with a sharp image of one person. A front-facing or slightly angled portrait, even lighting and an unobstructed face make the subject easier to interpret. Avoid heavy filters, sunglasses and group photos. The creator accepts JPG, PNG and WebP images up to 10 MB.
Use your own portrait or obtain explicit permission from the person pictured. A public image is not automatically permission to use someone’s likeness.
2. Watch the reference before uploading
Open the reference room and watch the full preview. Look at the entrance, the body movement, the cup and the camera framing. The footage is an original trend reference, not a generated result from this site.
The generator uses a silent reference. It does not bundle the original song or promise an exact copy of the scene.
3. Open the creator and choose your finish
On the video generator, choose your photo, then select 480p or 720p. Both produce a 15-second request in the current release.
| Setting | Credit cost | Consideration |
|---|---|---|
| 15 seconds · 480p | 100 credits | A lower-cost first attempt |
| 15 seconds · 720p | 300 credits | A higher-resolution finish |
Higher resolution does not guarantee better identity consistency or motion. If you are testing a new portrait, start with a setting that fits your budget.
4. Sign in and check your balance
You can prepare a local preview before signing in. The browser retains a portrait draft for up to an hour, so you can return to the creator after login. The photo is uploaded to the service when you start generation.
Review your balance and the quoted credit cost. If needed, compare the one-time credit packs. Checkout is available only when the payment service is configured; credits appear after a verified successful payment.
5. Confirm permission and submit once
Confirm that you have permission to use the photo and that you will not misrepresent the output as authentic footage. Select Generate video when the service is available. Your video request will appear in your task history.
Processing time depends on the provider and queue. If a request is still processing, check the existing task before submitting another one. A delayed task is not necessarily a failed task.
6. Review and save the complete clip
Watch the entire result. Check the face during turns, the hand holding the cup, clothing continuity, background changes and the camera movement. AI generation may distort details or change the subject’s appearance.
Open or save the result from your history. If you publish it, label it as AI-generated and add only music you have permission to use. Keep an honest distinction between an illustrative AI clip and real footage.
What if I want 28 seconds or a different scene?
The current creator supports 15 seconds with the fixed reference only. Longer videos, friend-group shots and custom scene options are not available. A separate editing or extension workflow would need its own quality checks.