Article
What Is Seedance 2.0?
Seedance 2.0 is ByteDance's latest multimodal AI video generation model, available through Jimeng (即梦) and integrated natively on HiArt. Unlike text-only video models, Seedance 2.0 accepts images, reference videos, and audio alongside your prompt — then generates a short clip with matching sound effects or background music.
If you have heard about "omni-reference" or the
@ syntax in AI video circles, that is Seedance 2.0's signature workflow: you upload assets, assign each one a role in natural language, and the model stitches motion, style, and rhythm together in a single 4–15 second render.What makes it different
- Multimodal inputs — combine up to 12 reference files (images, videos, audio) in one generation
- Built-in audio — outputs include synced sound effects or background music by default
- Director-level control — replicate camera moves, choreography, visual effects, and beat-matched motion from reference media
- Multiple aspect ratios — 16:9, 9:16, 1:1, 4:3, 3:4, 21:9, or adaptive for cross-platform exports
- Canvas-native on HiArt — iterate on the same node without leaving your project
Three generation modes
Seedance 2.0 supports three distinct workflows. Pick the one that matches your starting assets.
Text-to-video
Describe the scene, action, and camera in plain language. No uploads required — ideal for quick concept tests and mood boards.
First / last frame
Upload one image as the opening frame, or two images as the start and end points. The model generates the motion between them — useful for product reveals, logo animations, and controlled transitions.
Omni-reference
The most powerful mode. Upload reference images for character look or scene style, reference videos for camera movement or choreography, and reference audio for rhythm or mood. Use
@Image1, @Video1, @Audio1 syntax in your prompt to tell the model what each file is for.Input limits (omni-reference)
Input | Limit | Format | Max size |
|---|---|---|---|
Images | ≤ 9 | jpeg, png, webp, bmp, tiff, gif | 30 MB each |
Videos | ≤ 3 | mp4, mov | 50 MB each, 2–15s total |
Audio | ≤ 3 | mp3, wav | 15 MB each, ≤ 15s total |
Total | ≤ 12 files | — | — |
Output specs
- Duration: 4–15 seconds (selectable)
- Resolution: 480p or 720p
- Audio: auto-generated sound effects / background music
- Aspect ratio: 16:9, 9:16, 4:3, 3:4, 1:1, 21:9, or adaptive
Seedance 2.0 vs 2.0 Fast
HiArt offers two variants of the same model family:
- Seedance 2.0 — full-quality renders for final output. Use when you are happy with the prompt and ready to ship.
- Seedance 2.0 Fast — same multimodal controls with faster turnaround. Ideal for motion previews and prompt iteration before committing to a final render.
A typical workflow: draft with 2.0 Fast, refine your prompt and references, then switch to 2.0 for the final clip.
Who is it for?
- E-commerce & ads — product loops, hero shots, and 6-second social placements
- Short drama & storytelling — emotional close-ups, dialogue-driven scenes, beat-synced cuts
- Social content — dance replication, trend remixes, vertical 9:16 exports
- Creative prototyping — one-take tracking shots, VFX replication, style transfer from reference video
Things to know before you start
- Uploaded images and videos cannot contain realistic human faces — the platform blocks such content for compliance.
- Reference audio cannot be used alone; pair it with at least one reference image or video.
- First/last frame mode and omni-reference mode cannot be mixed in a single generation.
- Results depend heavily on prompt quality — especially how clearly you assign roles to each
@reference.
Ready to generate? Open a video node on HiArt canvas, select Seedance 2.0, and start with a simple text-to-video prompt. When you are ready to go deeper, read our Seedance 2.0 prompt writing guide for the
@ reference system, camera language, and copy-paste templates.