Seedance 2.5 Uncensored Video
Create 4–30s videos from a description, or steer motion using first and last frame images. Choose 480p or 720p output with optional synced audio.
Overview
Seedance 2.5 Uncensored is a video generation model by ByteDance. It turns a written scene description into a continuous video clip, and it can also animate from a starting image. You can guide the clip’s trajectory by supplying a first frame and an optional last frame, then the model fills in motion between them. This solves short-clip stitching for many workflows and helps you land on a planned opening and ending beat.
What you can build
- Text-to-video shots for ads, social posts, and b-roll
- Image-to-video animations from a single still frame
- Keyframed transitions that start on one image and end on another
- Reference-driven clips that keep a character, an outfit, or a product consistent through the shot
- Product or character clips that keep the same framing across a full take
- Cinematic camera-move experiments (push-in, orbit, handheld) driven by prompt
Inputs
- A written description of the scene and motion. This can include camera direction, lighting, style, and any dialogue or sound cues.
- An optional first-frame image you upload to anchor the opening frame. When you use a first frame, the output keeps that image’s aspect ratio.
- An optional last-frame image you upload to target the final frame. This only works when you also provide a first-frame image.
- An optional set of reference images, up to 30, that lock in a character’s face, an outfit, a product, or a visual style. Call them out in your description as [Image 1], [Image 2], and so on. Reference images replace first and last frame guidance rather than adding to it: when you supply them, the model composes the opening frame itself instead of starting from your image.
- A required output resolution choice. You can generate at 480p or 720p.
- An optional aspect ratio selection. You can pick common ratios or let the model decide. If you use a first-frame image, the image aspect ratio overrides this.
- A required duration selection from 4 to 30 seconds.
- A required choice to include synchronized audio or generate a silent clip.
- An optional output container choice. MP4 is the broad compatibility option. MOV is intended for higher color precision workflows.
- An optional watermark toggle.
Outputs
The model returns a single generated video file.
- Video format: MP4 or MOV, depending on your selection
- Visual specs: 480p or 720p, with the chosen duration and aspect ratio rules applied
- Audio: included only when you enable audio generation, and it is timed to the visuals
Recommended settings
- Provide a first frame when the opening framing and aspect ratio have to be exact.
- Reach for reference images instead when the subject matters more than the framing. Three to nine clean, consistent images of the same subject work best, and referring to them by number in the prompt keeps them straight.
Limitations
- Output is limited to 480p or 720p.
- Duration is limited to 4–30 seconds per clip.
- A last-frame target requires a first-frame image.
- Reference images cannot be combined with a first or last frame in the same generation.
- If your first/last frame images are low quality, heavily compressed, or inconsistent in style, the motion between frames can drift.
- Prompts that request exact replicas of real people, brands, or copyrighted characters may be blocked or degraded by provider-side rules.
Safety & compliance
- Get consent before generating video that depicts a real person, especially in sensitive contexts.
- Don’t use the model to create deceptive content (impersonation, fraud, or misleading political media).
- Respect copyright and trademark rules. Avoid prompts that request protected characters, logos, or scenes you don’t have rights to.
- “Uncensored” typically means fewer prompt blocks, not permission to break laws or platform rules. You are responsible for compliant use.
Example prompts
Great starting points for seedance-2.5-uncensored.
API quick start
Run seedance-2.5-uncensored with a single API call.
{
"prompt": "A lone surfer on a longboard carves throu…",
"resolution": "720p",
"duration": 5,
"generateAudio": "false"
}curl -X POST "https://api.wiro.ai/v1/Run/bytedance/seedance-2-5-uncensored" \
-H "Content-Type: application/json" \
-H "x-api-key: YOUR_WIRO_API_KEY" \
--data-binary @- <<'JSON'
{
"prompt": "A lone surfer on a longboard carves throu…",
"resolution": "720p",
"duration": 5,
"generateAudio": "false"
}
JSONOther models you might like
Explore similar video generation models.