Seedance 2.0 Face
Face-referenced Seedance 2.0: lock one person's likeness and generate new performances around them, scene after scene.
What is Seedance 2.0 Face?
The single hardest problem in AI video is keeping the same person looking like the same person. Seedance 2.0 Face solves it by treating a face reference as a hard constraint rather than a suggestion: upload a photo of your character (or actor), and the model generates new footage — new angles, new expressions, new lighting — while preserving the identity that reference defines.
This turns AI video from a novelty into a production tool. One reference portrait becomes a cast member who can appear in every scene of a series, every ad variant, every episode — with the same face every time. Dialogue-driven content, recurring hosts, and episodic mini-dramas are the obvious applications.
Key Capabilities
Identity lock
Facial structure, skin tone and distinguishing features stay anchored to your reference across the whole generation.
New performances
The model generates fresh expressions, head turns and actions — it does not merely animate the reference photo.
Scene flexibility
Change wardrobe, lighting, location and camera while the face stays recognizably the same person.
Series-ready
Reuse the same reference across projects to build a consistent recurring cast.
Prompt Examples
Copy any of these straight into the workspace.
[Face reference: uploaded portrait] The same woman, now in a bright kitchen, laughing while talking to camera and gesturing with a coffee mug. Natural daylight, handheld feel, vlog framing.
[Face reference: uploaded portrait] Profile shot of the same man walking through an airport terminal, pulling a suitcase, glancing at the departures board. Documentary style, soft window light.
What Seedance 2.0 Face Is Best For
- Episodic content with a recurring lead character
- Brand mascots or spokespersons that must match across campaigns
- Dialogue and reaction scenes where the face carries the emotion
- Virtual influencer and avatar-first formats
120s
Est. Latency
Seedance
Model Vendor
1080P HD
Max Quality
Seedance 2.0 Face FAQ
How do I use a face reference with this model?
Upload a clear, well-lit photo of the person in the video workspace, select Seedance 2.0 Face, and write your prompt around "the same person". The model treats the reference as the identity anchor for the whole generation.
Will the face match my reference exactly?
It preserves identity — the recognizable structure and features — while rendering new expressions, angles and lighting. It will not reproduce the photo pixel-for-pixel, because it is generating a performance, not editing an image.
What is the difference between Seedance 2.0 Face and Seedance 2.0 Fast Face?
They share the same face-anchoring technology. Face is the fuller-quality render; Fast Face completes in roughly two-thirds of the time and is better for iteration and drafts.
Can I use the same face across multiple videos?
Yes — that is the core use case. Keep the same reference image and reuse it in every project; the model will cast the same "actor" each time.
Explore Related Models
View allSeedance 2.0
Fluid, cohesive multi-shot video outputs
Seedance 2.0 Fast
Fast high-quality video generation, rich frames, first choice for quick iterations
Seedance 2.0 Fast Face
Fast face-referenced driven, lightweight and efficient character generation
Seedance 1.5 Pro
Precise audio-video synchronization, cinematic camera movement, supports background music
Ready to Create with Seedance 2.0 Face?
Every model is included in your plan and billed in credits — no setup, no API keys, generation starts in seconds.