AI video generator
Pick a Seedance tier, choose how the clip starts, set duration, resolution and aspect ratio, then generate.
Results
Generated content will appear here
The four Seedance tiers
Each card opens the page for one tier, with its own sample clips, published prompts and limits. All four run in the workspace on this page.
AI image generator
Generate images with Nano Banana 2.1, Nano Banana Pro, Nano Banana 2, Nano Banana 2 Lite, GPT Image 2.5, GPT Image 2 and FLUX 3 Image in one workspace.
Create video with Google Veo 3.1
Google Veo 3.1 generates 4, 6 or 8 second clips with native audio. Official specs, prompts published by Google, and what this workspace exposes.
Create video with Happy Horse 1.1
Happy Horse 1.1 by Alibaba generates 3-15 second videos from text, an image or references. Official specs, a published prompt and what this workspace exposes.
Seedance 2.5 video generator with reference control
ByteDance Seed's Seedance 2.5 generates up to 30 seconds in one take, with reference control. Official specs, published prompts and this workspace's limits.
Seedance 2.0 video generation with multimodal references
Seedance 2.0 takes text, image, video and audio references and returns a 15-second multi-shot clip with dual-channel audio. Official specs, eight sample clips.
Seedance 2.0 Fast video generator
Seedance 2.0 Fast is the middle tier of ByteDance’s 2.0 family: 4-15 second clips at 480p or 720p, native audio, no 1080p. Official rows, published prompts.
Seedance 2.0 Mini video generator with the official model card
Seedance 2.0 Mini is ByteDance’s cheapest Seedance 2.0 tier: official model ID, 480p and 720p at 4 to 15 seconds, per-tier prices and named clips.
One workspace, three ways to start a clip
This page is the video workspace itself. A clip can start from a written shot, from an image you already have, or from reference material, and the tier you pick decides which of those a run accepts.
- Text to video: describe the shot, including camera movement and lighting, and generate it.
- Image to video: attach a still and describe how it should move.
- Reference-guided: attach the material the clip has to follow, such as a look or a motion pattern you want carried across.
The four tiers side by side
All four tiers sit in the model picker. The differences that change what you can actually ask for are the clip length, the resolution tiers and how many images an image-to-video run takes.
| Tier | Clip length | Resolution tiers | Images per image-to-video run |
|---|---|---|---|
| Seedance 2.5 | 4–30 s | 480p, 720p, 1080p | up to 2 |
| Seedance 2.0 | 2–15 s | 480p, 720p, 1080p | 1 |
| Seedance 2.0 Fast | 2–15 s | 480p, 720p | 1 |
| Seedance 2.0 Mini | 2–15 s | 720p, 1080p | 1 |
How a run works
Five steps, in the order the form asks for them.
- Pick a tier. The picker decides which clip lengths, resolution tiers and inputs are on offer.
- Choose how the clip starts, then supply whatever that mode needs: a written shot, a still, or reference material.
- Write the shot as a sequence of moments rather than a list of nouns. Naming the camera move and the light does more work than piling up adjectives.
- Check the price the workspace quotes, then send the run.
- Watch the whole clip before you publish it. Motion artefacts usually land in fast movement, in hands, and in whatever the camera passes behind.
What moves when you change tier
The tiers are not simply better and worse versions of each other. Each one changes a different constraint, so the useful question is which constraint is binding for the clip you have in mind.
- Seedance 2.5 reaches 30 seconds in a single generation, which is the one thing the 2.0 tiers cannot do. It also takes two images in image-to-video where they take one.
- Seedance 2.0 adds 1080p to the same 2-to-15-second range as Fast.
- Seedance 2.0 Fast caps at 720p, which is often enough for a draft you intend to regenerate.
- Seedance 2.0 Mini runs at 720p or 1080p but offers the shortest aspect-ratio list of the four, so check that the format you need is in the picker before you commit.
What this page does not decide
Prices, parameters and the tier line-up come from the live service. This page describes what the workspace offers today, and the workspace wins if the two ever disagree.
- The point cost of a run is quoted by the service before you send it, not estimated here. A longer clip and a higher resolution tier are both charged differently, so trust the figure in the form rather than a figure on a page.
- A tier can be added or retired without this page being rewritten. The picker is the authority.
- The clips below are their publishers' own samples. Nothing on this page grants a licence to reuse them.
Examples
Seedance 2.5, the tier that reaches 30 seconds in one generation. This is the official narrative sample: 28 seconds in a continuous take, with the published prompt opening on the room and the light before anything moves.
Seedance 2.0 at 16:9, from the official gallery. The prompt asks for butterfly lighting and a cut into close-up, which is the kind of staging the 2-to-15-second range is built for.
Seedance 2.0 Mini, the lightest of the four. A third-party capability clip published under a block about fast drafts for high-volume creative testing, which is the job this tier is positioned for.
Questions
Which video tiers can I use here?
Four: Seedance 2.5, Seedance 2.0, Seedance 2.0 Fast and Seedance 2.0 Mini. All four are in the model picker on this page.
How long can a clip be?
Seedance 2.5 generates between 4 and 30 seconds in one run. The three Seedance 2.0 tiers generate between 2 and 15 seconds. The form sets the exact length inside that range.
Can I animate a photo I already have?
Yes. Choose image to video, attach the still and describe how it should move. Seedance 2.5 accepts up to two images in that mode; the three 2.0 tiers accept one.
Which tier should I start with?
Seedance 2.5 when the clip has to run long as a single take, or when you want the widest choice of formats. Move to a 2.0 tier for a short draft, for a lower resolution while you iterate, or when the run has to stay light.
Do I see the clip before I pay for it?
No. The form quotes the price before the run is sent, and the clip is generated once you send it. Generating a batch of variations and choosing between them afterwards is not how this workspace is billed.
Do the clips on this page come from this site?
No. Each clip below is a sample its publisher released, credited in full underneath. They show what each tier produces, and none of them was generated here.
Why is there no price table on this page?
Because a run's cost moves with the tier, the clip length, the resolution and the mode, and a written table would go stale without anyone noticing. The form quotes the live figure instead.