How to Use Seedance 2.0: Where to Access It and How to Make Your First Video
How to use Seedance 2.0: where to get access in the US and elsewhere, which mode to pick, how to write the prompt and which settings to choose.
By Melies AI · Updated · 7 min read

Seedance 2.0 is ByteDance’s AI video model for clips of 4 to 15 seconds with sound. You cannot download it, so you use it through an app or website that runs it: ByteDance’s own Dreamina and CapCut, or independent platforms such as ElevenLabs and Higgsfield. This guide covers where to get access, how each mode works, how to write a prompt, and what to do when a clip is not right. For background on the model itself, see every Seedance version and its release date.
Where can I use Seedance 2.0?
You can use Seedance 2.0 in ByteDance’s Dreamina and CapCut apps, in the ElevenLabs ElevenCreative workspace, on Higgsfield, through ByteDance’s BytePlus API, and on browser sites such as Melies AI. In China, ByteDance runs it in the Jimeng AI and Doubao apps. The platforms differ in how you pay, what you get for free and who they suit.
| Platform | Run by | How you get access | Free use | Suits |
|---|---|---|---|---|
| Dreamina | ByteDance | Web and app, monthly plans | 120 credits a day on US free accounts; free downloads carry a watermark | ByteDance’s own creative app |
| CapCut | ByteDance | AI video tools inside the CapCut editor | Depends on region and plan | People who already edit in CapCut |
| ElevenLabs | ElevenLabs | ElevenCreative workspace, paid plans only | Not on the free plan | Adding voice-over, music and sound effects in one place |
| Higgsfield | Higgsfield | Monthly plans with credits | No free plan listed | Many video models under one plan, up to 4K |
| BytePlus ModelArk | ByteDance | API for developers, prepaid packs or pay as you go | Not listed | Building Seedance into your own app; not available in the US |
| Melies AI | Independent | Browser, one-time credit packs | 4 credits on Google sign-up | Paying per clip with no subscription |
ByteDance Seed’s own Seedance pages also link to Dola, ByteDance’s AI assistant app, as a place to try the model; what Dola offers depends on your region. Plans and free allowances change often, so check the platform before you pay. This table was checked on November 18, 2026, and it leaves out prices on purpose.
Is Seedance 2.0 available in the US?
Yes. People in the United States can use Seedance 2.0 on Dreamina, CapCut, ElevenLabs, Higgsfield and browser sites such as Melies AI. Dreamina gives US free accounts 120 credits a day, and on August 6, 2026 Dreamina announced that Seedance 2.5 had also gone live for US users. The one exception is ByteDance’s developer cloud: BytePlus states that it is not available in the United States, so US developers reach the Seedance API through other providers.
How do you make a video with Seedance 2.0, step by step?
Making a Seedance 2.0 video takes six steps on any platform: pick the model, choose a mode, write the prompt, set the length and quality, generate, then refine. The menus have different names on each site, but the settings are the same model settings underneath.
- 01
Pick a platform and select Seedance 2.0
Open an app or website that runs Seedance 2.0, such as Dreamina, CapCut, ElevenLabs, Higgsfield or a browser tool, and choose Seedance 2.0 in its model menu. Seedance 2.0 Fast and Seedance 2.0 Mini are lighter versions for quicker or cheaper drafts.
- 02
Choose a mode
Text to video starts from a prompt alone. Image to video animates a first frame, with an optional last frame. Reference to video takes several images, and on some platforms video and audio clips too, which you name @Image1, @Image2 and so on in the prompt.
- 03
Write the prompt
Describe the subject, the action, the setting, the camera move, the style and the sound, in that order. Put spoken lines in quotes. For several shots, label them “Shot 1”, “Shot 2” or give time ranges such as “0–5s”.
- 04
Set length, shape and quality
Seedance 2.0 makes 4- to 15-second clips. Pick the aspect ratio for where the video will be posted, 480p for drafts and 720p or 1080p for the final, and switch sound on or off where the platform offers it.
- 05
Generate and download
Press Generate and wait a few minutes. Watch the result, change one thing in the prompt if it is not right, and download the MP4. Free plans on some platforms add a watermark to downloads.
- 06
Extend or edit the clip
ByteDance built Seedance 2.0 to extend a clip and to edit parts of it from a new prompt. Not every platform shows these tools, so check the platform you use; otherwise generate a follow-up clip from the last frame.
Which mode should I pick: text, image or reference?
Pick text to video when you only have an idea, image to video when the first frame must look exactly like your picture, and reference to video when characters, products or a location must match your images across a whole clip. ByteDance’s Seedance 2.0 accepts up to 9 images, 3 video clips and 3 audio clips as references in one generation; some platforms, including Melies AI, take images only.
| Mode | Starts from | Best for | Tip |
|---|---|---|---|
| Text to video | A prompt | Ideas, scenes and shots you can describe | Name the camera move and the sound |
| Image to video | A first frame, plus an optional last frame | Animating a photo, product shot or artwork | Describe only the motion, not what is already in the image |
| Reference to video | Up to 9 images (video and audio on some platforms) | Keeping a character, outfit or product the same | Say what each reference is for: “@Image1 is the main character” |
How do I prompt Seedance 2.0?
A good Seedance 2.0 prompt reads like a short shot list: subject, action, setting, camera, style and sound, in that order. One clear action per shot works better than a long list. Put dialogue in quotes so the model speaks it, and label several shots “Shot 1”, “Shot 2” or with time ranges. ByteDance’s launch post shows prompts in this form, such as this image-to-video example:
A girl hangs laundry gracefully. After finishing, she takes another piece of clothing from the bucket and shakes it vigorously.
The same structure, written for a multi-shot text-to-video clip with sound:
Shot 1 (0–5s): Wide shot of a night market in the rain, neon signs reflected in puddles, slow push-in. Shot 2 (5–10s): Close-up of a street cook flipping noodles in a wok, flames rising, handheld camera. Shot 3 (10–15s): She looks up and says, "Last bowl tonight!" Sound: sizzling wok, rain, distant chatter.
- Keep the style in one line. “35mm film, warm tungsten light” beats a list of ten adjectives.
- In image to video, describe motion only. The picture already sets the look, so the prompt should say what moves and how the camera moves.
- Real people need permission. ByteDance requires identity verification or legal authorization to use real human portraits as references, so platforms often refuse photos of real faces.
How do I use Seedance 2.0 for free?
You can try Seedance 2.0 for free with a small daily or one-time allowance, but no platform offers unlimited free Seedance 2.0 video. Dreamina gives US free accounts 120 credits a day, with a watermark on free downloads. Melies AI gives 4 credits on Google sign-up, enough for one 4-second Seedance 2.0 Mini clip at 480p with no watermark; try Seedance 2.0 in the browser to use it. ElevenLabs offers the model on paid plans only, and Higgsfield lists no free plan.
Can I download Seedance 2.0 or run it locally?
No. ByteDance has not released Seedance 2.0’s model weights, so there is no file to download and no way to run it on your own GPU or in a local ComfyUI install. “Downloading Seedance” in practice means installing an app that runs it, such as CapCut or Dreamina, or using a website. Developers can call it through an API instead.
Should I use Seedance 2.0 or Seedance 2.5?
Use Seedance 2.0 for clips of 15 seconds or less: on Melies AI it costs less per second than Seedance 2.5 at every quality. Use Seedance 2.5 when one take must run longer: ByteDance released it on July 31, 2026 with clips of up to 30 seconds, more reference inputs (up to 30 images, 10 video clips and 10 audio clips) and timestamp-level editing. On Melies AI, Seedance 2.5’s 30-second clips run at up to 30 reference images, against 9 on Seedance 2.0.
Seedance 2.0 on Melies AI makes 4- to 15-second clips in 16:9, 9:16, 1:1, 4:3, 3:4 and 21:9, or adaptive, from 480p to 1080p. To compare what each clip costs, see Seedance pricing per second.
Sources
- ByteDance Seed: Seedance 2.0 official launch
- ByteDance Seed: Seedance 2.0 technical report
- ByteDance Seed: Introducing Seedance 2.5
- BytePlus: Dreamina Seedance 2.5 is now available on BytePlus
- Dreamina: free credits, limits and exports
- Dreamina on X: Seedance 2.5 live in the US (Aug 6, 2026)
- ElevenLabs: Seedance 2.0 in ElevenCreative
- Higgsfield: Seedance 2.0 pricing and platforms (Oct 2026)
- CapCut: Dreamina Seedance 2.5
