Skip to main content

Put it into words. Create what you imagine

Generate video with Seedance 2.0 and 2.5, or images with Seedream, from text or an image

Highlights

Showcase

Product Cinema

Campaign Energy

Work, Reimagined

Close to Heart

Micro Stories

High Stakes

Animated Worlds

Culture in Motion

Beyond Reality

VIDEO · IMAGE · DRAMA

Choose the first frame, add more, join several clips into one

Video runs on Seedance 2.0 and 2.5, images on Seedream 5.0. Video and image generation shows the credits it will use before it runs.

A service that makes video and images from text or images. Video runs on Seedance 2.0 and 2.5, up to 30 seconds on 2.5; images on Seedream 5.0, up to four per turn. Videos can be extended, changed and joined, and there is AI drama production too.

Showcase

Product Cinema

E-commerce and consumer products

Campaign Energy

Advertising and brand films

Work, Reimagined

B2B, industrial and services

Close to Heart

Romance, longing and connection

Micro Stories

Contemporary drama, comedy and everyday tension

Canvas

Connect images, video, audio and prompts to generate. Earlier results stay when you branch.

AI Drama

Organize characters and shots from your story, then produce the video.

Video generation

Attach references and specify camera, light and motion. Seedance 2.5 supports dialogue-driven lip movement, up to 30 seconds.

Image generation

Specify composition, light and texture, then choose the image to use.

Performance Remake

Use Skill, animation, CG or AI videos. Live-action video of real people is not supported yet.

Canvas Trace each take back along its line

Connect images, video, audio and prompts to generate. Earlier results stay when you branch.

Open Canvas

AI Drama From synopsis to storyboard

Organize characters and shots from your story, then produce the video.

Open AI Drama

Video generation From text or from an image

Attach references and specify camera, light and motion. Seedance 2.5 supports dialogue-driven lip movement, up to 30 seconds.

Open Video generation

Image generation Up to four a turn, and a first frame for video

Specify composition, light and texture, then choose the image to use.

Open Image generation

Performance Remake Reference the original motion, change characters, clothing and setting

Use Skill, animation, CG or AI videos. Live-action video of real people is not supported yet.

Open Performance Remake

BUSINESS

For companies

Commercial use and rights: what to check before using it for work. Summarised from the Terms.

Commercial use

No clause prohibits use in advertising or promotion. Before publishing, check that it infringes no one else’s rights (Article 8). Hiding that something was made with AI in order to deceive people is prohibited (Article 10).

Rights and likeness

Between us and you, any rights that arise in the video and images you make belong to you. We do not guarantee that rights arise. Using a real person needs that person’s consent, or a legal guardian’s for a minor (Articles 8 and 9).

Start from what it’s for

Short drama

A three-shot prompt: establish the scene, exchange dialogue, show the reaction.

Start with this

Product spot

A two- or three-shot product prompt: use, material close-up, still display. No on-screen text.

Start with this

Fan MV

Follow one lead. An attached track prompts cuts on its beats.

Start with this

VTuber short

One take to camera, with a fixed camera and one gesture per shot.

Start with this

Ads and brand films

When the mood matters more than the product. Describe the scene and it is made into one video as written.

Start with this

Images for posts and announcements

Make up to four a turn and pick the one for your post or announcement.

Start with this

How it differs from other ways

Add more

Shoot and edit
Shoot the next scene and join it in the edit
One-off generation
Describe the next scene and make it as a new clip
Studio
Generate more before or after the clip you made

Change one thing

Shoot and edit
Reshoot, or replace it in compositing
One-off generation
Rewrite the prompt and remake the whole clip
Studio
Describe the part to change and ask for it to be replaced

Join into one

Shoot and edit
Line clips up in an editor and smooth the cuts
One-off generation
Export the clips and join them in an editor
Studio
Join two or three clips, describing how the cuts look

Image to video

Shoot and edit
Animate from a drawing or a photo
One-off generation
Hand over an image and make a clip from it
Studio
The image you pick becomes the first frame and moves from there

Restyle

Shoot and edit
Reshoot, or redraw it with CG and compositing
One-off generation
Remade from text, so the motion changes too
Studio
Reference the original motion, change characters, clothing and setting

Synopsis to storyboard

Shoot and edit
Write the script and draw the storyboard by hand
One-off generation
Write a prompt per scene and make each clip in turn
Studio
Organize characters and shots from your story, then produce the video.

Video with sound, and images, in one workspace

VideoSeedance 2.0 / Seedance 2.5

Make video, then fix it

Make 4–15 second clips (up to 30 on Seedance 2.5) from text or an image. Fix a finished clip by extending it, changing one part or joining two or three clips.

ImageSeedream 5.0 / Seedream 5.0 Pro

Make images

Up to four per turn, from text or reference images. Seedream 5.0 goes up to 4K, 5.0 Pro up to 2K. An image you make can be a video’s first frame.

Audioaudio on

Video with sound

Turn sound on or off for each generation.

What it does, in numbers

30s

The longest Seedance 2.5 clip. It can come with sound that follows the picture.

4 images

Images from one prompt. Pick the one you use.

8 ways

From text, image to video, add more, change one thing, join into one, restyle: the video operations.

Make your first one

Questions

How does pricing work?

Generating uses credits, which you buy in advance. Each credit product’s price and credit count are on the pricing page.

See pricing
How many credits does one generation use?

It depends on the model, the length, the resolution and more. The number is shown before you generate. If a generation does not finish, the credits held for it are not spent and come back (Terms, Article 5).

Rights, and real people

Between us and you, rights in the output belong to the member who made it (Terms, Article 8). We do not guarantee that rights arise, or that it infringes no one else’s rights. A real person may appear only with their consent (a legal guardian’s for a minor; Article 9).

Can I use it for work or advertising?

The Terms do not prohibit commercial use. Before publishing, check that the output infringes no one’s rights. Hiding that it was AI-generated in order to deceive people is prohibited (Article 10).

Do credits expire?

No. If you close your account, unused credits lapse and, except where the law requires otherwise, are not refunded (Terms, Articles 5 and 15).

Can I try it without signing in?

While the trial is open, you can make video without signing in, with Seedance 2.0, 2.0 Fast and 2.0 Mini. Trial videos carry a moving watermark, and their sound stops halfway. Seedance 2.5 and buying credits need you to sign in. There are periods when the trial is not open, and days when that day’s trial allowance has run out.

Which models can I use?

Four for video: Seedance 2.5 makes 4–30 seconds up to 1080p, 2.0 makes 4–15 seconds up to 4K, and 2.0 Fast and 2.0 Mini make 4–15 seconds up to 720p. Two for images: Seedream 5.0 Pro (up to 2K) and Seedream 5.0 (up to 4K). The models on offer can change (Terms, Article 13).

Can I continue a video I made?

Yes. Add more puts new footage after the video or before it. You can also change one part, or join two or three clips into one. Source videos can total up to 30 seconds with Seedance 2.5 and up to 15 seconds with the 2.0 models.

Which languages does it support?

The interface supports Japanese, English, Simplified Chinese and Korean. Switch in the language menu or workspace settings. Shape it follows the language of your input.

How are my inputs and work handled?

To generate, your inputs are sent to outside AI providers (Terms, Article 8). Your inputs and outputs are not used to train AI models without your consent (Privacy Policy, section 3). Your work goes into the public gallery or advertising only with your consent (Terms, Article 8). You can delete images and videos you uploaded, and videos and images you generated, from Assets. A file that a project or similar item uses can be deleted once you remove it there, where that is possible. A generation that used or made a deleted file stays in your generation history with its prompt and other records. A deleted file can no longer be used in the service and is kept as an archive. To have deleted what you cannot delete yourself, such as a file in a published work, or to have the archive erased, ask us through the contact in the Privacy Policy, and we handle the request as the law requires (Privacy Policy, sections 9 and 10). To close your account, write to the contact in Privacy Policy section 11 (Terms, Article 15).