Seedance 2.0 · Reference to Video

ABOUT THE WORKFLOW

Combine References into Video
Upload a reference image of the subject you want in the video, then upload a reference video showing the motion or action you want them to perform. Seedance 2.0 reads the identity from the image (face, clothing, build) and the motion from the video (action, camera, pacing), then generates a new video that puts your subject into the reference motion. You can add up to 9 images, 3 videos, and 3 audio references in a single generation. Audio is generated natively alongside the video.

Partner node. This workflow calls an external API, so each run uses credits from your API wallet. No API key needed. Floyo handles the connection.

Model

• Seedance 2.0 by ByteDance (SEED Lab). The #1-ranked video model on the Artificial Analysis Video Arena for both text-to-video and image-to-video. Built on a unified multimodal audio-video architecture that accepts text, images, video, and audio references simultaneously. Generates clips from 4 to 15 seconds with native audio at up to 1080p.

HOW IT WORKS

Step 1. Upload your character reference image
A clear photo of the person, character, or subject you want in the video. The model preserves face, clothing, accessories, and proportions from this image.
Works great with: portraits · character renders · product shots · AI-generated characters

Step 2. Upload your motion reference video
A video showing the action, camera movement, or performance you want your character to reproduce. The model reads the motion, timing, and camera language from this clip.

Step 3. Write your prompt
Describe how the references combine. The default is "Use the person in the reference image as the main subject, and recreate the reference video." Edit this to add specific direction or constraints.

Step 4. Add more references (optional)
Upload additional images (up to 9 total), videos (up to 3), or audio files (up to 3) to guide the generation. More references give the model more context for style, environment, sound, and secondary characters.

Step 5. Hit run and download
Seedance 2.0 generates the video with native audio and returns an MP4.
Ready for: Premiere · DaVinci Resolve · After Effects · TikTok · Instagram · YouTube

First time? Upload one character image, one motion reference video, and hit run. Leave the prompt and settings as-is.

RECOMMENDED SETTINGS

Quick-start guide. Find the goal that matches yours and copy the settings.

• Standard motion transfer (most people) — One character image + one motion video, 480p, 4 seconds, audio on. No changes needed.

• Higher resolution — Switch from 480p to 720p or 1080p. Higher resolution uses more credits and takes longer, but produces sharper output.

• Longer clip — Increase duration from 4 seconds up to 15 seconds. Longer clips give the model more time to develop the referenced motion and camera work.

• Different aspect ratio — Switch from 16:9 to 9:16 for vertical social content, 1:1 for square, or 21:9 for ultrawide cinematic framing.

• Multiple characters — Upload separate character reference images in slots 1 through 9. Describe each in the prompt: "The person from image 1 is the lead, the person from image 2 enters from the right." Seedance 2.0 handles multi-character scenes with distinct identity preservation.

• Audio reference — Upload an audio file as a reference to guide the soundtrack, rhythm, or voice. The model syncs generated audio to both the visual motion and the audio reference.

• Character identity drifts — Use a clearer, well-lit, front-facing reference image. Add identity preservation language to the prompt: "Preserve the exact face, hair, and outfit from the reference image."

Prompt: Start with what maps to what: "Use the person from image 1 as the subject and recreate the motion from video 1." Then add context: scene details, lighting, mood, or audio direction. The model reads references directly, so the prompt steers how they combine rather than describing everything from scratch.

LEARN

📹 Videos

• Intro to Floyo

• ComfyUI 101 Free Course ft. Sebastian Kamph

• Floyo 101 for Team Collaboration

✨ Quick links

• Help docs

• LoRA Training guide

• Discover more workflows

USE CASES

🎬 Motion Transfer for Film and Ads
Film a stunt or action sequence with stand-ins, then swap in your character using a reference image. The model reproduces the motion, camera work, and pacing with the new subject's identity preserved.

👤 Virtual Influencer and Character Content
Generate consistent video content of a character or virtual persona performing different actions. Upload the same character image with different motion reference videos to produce a library of clips.

🛍️ Product and Brand Videos
Reference a demo video showing a product interaction, then regenerate it with different subjects, settings, or styling while preserving the action choreography.

📱 Social Content at Scale
Upload one character and multiple short motion references to produce a batch of social clips with consistent identity and varied action for TikTok, Reels, and Shorts.

🎤 Music and Performance Videos
Reference a dance or performance video and transfer the movement onto a character or avatar. The model syncs motion with generated audio or an uploaded audio reference.

WHAT WORKS BEST / WHAT TO AVOID

✅ Works great

• Clear, well-lit character reference images (front-facing, visible face)

• Motion reference videos with clean, distinct action

• Prompts that specify how references combine ("person from image 1, motion from video 1")

• Single-subject motion transfer with moderate complexity

⚠️ May produce softer results

• Blurry or heavily occluded character reference images

• Very fast, chaotic motion in the reference video

• Reference videos with many overlapping subjects

• Conflicting visual styles between the character image and the motion video

FAQ

What is reference-to-video in Seedance 2.0?
Seedance 2.0 supports multimodal all-round reference, allowing combined input of various texts, images, videos, and audio. The model can accurately understand multimodal input content and generate output as instructed by referencing elements including visual composition, camera language, motion rhythm, and sound characteristics from the inputs. In this workflow, you upload a character image as the identity reference and a video as the motion reference. Higgsfield

How many references can I use at once?
This workflow supports up to 9 images, 3 videos, and 3 audio files in a single generation. Seedance 2.0 can generate clips from 4 to 15 seconds, supports up to 1080p, and works across multiple aspect ratios including 16:9, 9:16, 4:3, 3:4, 21:9, and 1:1. Synthesia

Does Seedance 2.0 maintain character identity across the video?
Yes. Character consistency remains one of the hardest parts of AI video generation. Seedance 2.0 is built to hold faces, clothing, accessories, and small subject details more consistently across the duration of a clip. Use a clear, well-lit reference image to give the model the strongest identity anchor. Synthesia

How does motion reference work?
The model reads the action, camera movement, timing, and pacing from your reference video, then reproduces those elements with the identity from your reference image. It is not a frame-by-frame copy. The model interprets the motion and regenerates it with the new subject, adjusting for differences in body type and proportion.

Does Seedance 2.0 generate audio?
Yes. Audio is generated natively in the same pass as the video. The model produces synchronized sound effects, ambient audio, and dialogue that match the on-screen action. You can also upload an audio reference to guide the soundtrack or voice.

Is Seedance 2.0 licensed for commercial use?
Seedance 2.0 is a proprietary ByteDance model. Commercial use is governed by ByteDance's terms of service. Outputs carry a C2PA provenance watermark embedded in the file metadata. Review the current terms and regional availability for your specific use case.

How to run Seedance 2.0 reference to video online?
You can run Seedance 2.0 reference to video online through Floyo. No installation, no setup, no API key to wire up. Open the workflow in your browser, upload your references, and hit run. Free to try.

WHY FLOYO?

Floyo is the only platform with team collaboration for ComfyUI in the browser. You run workflows with no install. You share run history, assets, and models across your team. You pay only when you generate. Floyo supports open-source and closed-source models.

A designer runs an edit and likes the result. A teammate opens that exact run from shared history and keeps going. No file handoffs. No version confusion.

For studios and enterprise teams, Floyo adds private workspaces, pooled resources, and a team usage dashboard. Other ComfyUI cloud tools run for one person at a time. Floyo runs for the whole team, with transparent per-generation costs.

Ready to try it?
Upload a character image, upload a motion reference video, and hit run.

→ Launch Workflow, Free

Questions? Watch the free course or check the FAQ above.