How to use Seedance 2.0 in one line: open Dreamina, pick Seedance 2.0 in AI Video, add your images, clips or audio, describe the shot and click Generate. Seedance 2.0 is ByteDance's video model. It turns text, pictures, video clips and audio into short clips that come with their own sound. (Dreamina: How to use Seedance 2.0 (opens in a new tab); ByteDance Seed: Seedance 2.0 official launch (opens in a new tab))

ByteDance's Seed team launched Seedance 2.0 on February 12, 2026. According to the launch post, one request can combine up to 9 images, 3 video clips and 3 audio clips with a written instruction, and the model returns up to 15 seconds of multi-shot video. The sound is generated with the picture: background music, sound effects and voices, in stereo. (ByteDance Seed: Seedance 2.0 official launch (opens in a new tab))

You do not need a developer account. ByteDance puts the model inside its consumer apps: Dreamina, its AI image and video site, and CapCut, where CapCut's newsroom says it rolled out in AI Video and Video Studio and reached users in the US on April 7. (Dreamina: How to use Seedance 2.0 (opens in a new tab); CapCut Newsroom: Dreamina Seedance 2.0 (opens in a new tab))

How to use Seedance 2.0 for free: what the credits get you

Partly. Dreamina gives free credits, and its own guide says the published August 2026 baseline was 120 credits a day, while warning that the amount can change and should be checked in your account. Each video costs credits, and Dreamina says the cost depends on the model, the length, the resolution, the mode and the extra features you turn on. It does not publish one fixed price per clip. (Dreamina: free credits and paid plans explained (opens in a new tab))

Dreamina also offers a lighter version, listed as Seedance 2.0 Lite on one page and Seedance 2.0 mini in the model selector, and says it can be tried with free trial credits. Dreamina now lists a newer Seedance 2.5 as well, and says free access to it varies by account. On CapCut, No Film School reported that new users get some free credits in Video Studio before paid pricing applies. (Dreamina: Seedance 2.0 Lite (opens in a new tab); Dreamina: Seedance 2.5 free online (opens in a new tab); No Film School: Seedance 2.0 in CapCut (opens in a new tab))

So free is real, but it is a trial budget, not an unlimited tool. Plan a few strong attempts a day instead of dozens of quick ones, and check the credit cost shown before you click Generate.

Two catches on the free side. Dreamina says a free generation does not guarantee a download without a watermark or at a given resolution; paid members may export without visible branding where the plan and model allow it. CapCut says it puts visible watermarks and C2PA Content Credentials on AI-generated content. (Dreamina: free credits and paid plans explained (opens in a new tab); CapCut Newsroom: Dreamina Seedance 2.0 (opens in a new tab))

How to use Seedance 2.0 in Dreamina, step by step

1. Sign in and open AI Video

Go to Dreamina and log in. Open AI Video and choose Seedance 2.0 as the model. If you only see a Lite or mini version, that is the cheaper model; the steps are the same.

2. Choose how you give it pictures

Dreamina's guide describes two modes. Single-frame lets you upload a first frame and a last frame, and the model fills in the motion between them. Multiframes lets you upload a mix of images, audio files and video clips as references.

3. Add your references

Upload only what matters for the shot: a character, a product, a location, a voice or a piece of music. Dreamina says video and audio references can be up to 15 seconds each, and the model takes up to 9 images, 3 videos and 3 audio clips.

4. Write the prompt and name your files

Describe the scene, what each character does and what you want to hear. When you use several files, refer to each one in the prompt by its name with an @ sign, the way Dreamina's guide shows (@AssetName). Then click Generate.

5. Review, fix and download

Watch the result. Click Regenerate if it missed. Dreamina has two one-click extras: Generate soundtrack adds audio, and Interpolate frames smooths the motion. When it looks right, click Download in the top right corner to save the 1080p video.

Dreamina's prompt advice is concrete: be specific about the composition of the scene, the action and the audio cues, describe small movements and expressions such as a slight head turn or natural blinking, and preview before you export to check timing, transitions and sound sync. (Dreamina: How to use Seedance 2.0 (opens in a new tab))

Prompts to copy

Prompt 1: Product shot from one photo

@Product sits on a wooden kitchen counter in soft morning light. The camera slowly pushes in and circles to the left. Steam rises from a cup beside it. Quiet acoustic guitar music, birds outside the window.

Prompt 2: Character with a voice

@Character stands on a rainy city street at night and looks up at the sky, then smiles and blinks naturally. Use @Voice for one short line of dialogue. Sound of rain and distant traffic. Three shots: wide, medium, close-up.

Prompt 3: First frame to last frame

Start on the first frame and end exactly on the last frame. The camera moves smoothly between them with no cuts. Keep the lighting and the colors the same the whole time. Soft ambient sound only.

Tip: Write what the camera does and what the viewer hears in separate sentences. Prompts that mix them in one long line are harder to fix when one part goes wrong.

Rules about real faces

Real people are restricted. ByteDance says using a real person's portrait as a subject reference needs identity verification or prior legal authorization. In CapCut's first rollout, the company blocked making videos from images or videos that contain real faces, and it says it blocks prompts and material that break its policies, including unauthorized use of intellectual property. (ByteDance Seed: Seedance 2.0 official launch (opens in a new tab); CapCut Newsroom: Dreamina Seedance 2.0 (opens in a new tab))

Tip: If your upload is rejected, check whether it shows a real face or a famous character before you rewrite the prompt. That is the most common reason a reference is refused.

Where Seedance 2.0 falls short

ByteDance is frank about it. The launch post says the model is still far from perfect: detail stability, realism and natural motion need work, keeping several characters consistent is hard, text inside the video often comes out wrong, and audio can occasionally distort. (ByteDance Seed: Seedance 2.0 official launch (opens in a new tab))

In practice that means short, simple shots with one main subject work best. Do not ask for readable signs or captions inside the clip; add text later in an editor. Long scenes with three or four people are where a free credit budget disappears fastest.

If you edit in CapCut anyway, Video Studio on CapCut Web is the other way in: No Film School describes it as a canvas-based workspace with no timeline, and CapCut says clips there can run up to 15 seconds in six aspect ratios. (No Film School: Seedance 2.0 in CapCut (opens in a new tab); CapCut Newsroom: Dreamina Seedance 2.0 (opens in a new tab))