ByteDance Seed · Seedance 2.5

Seedance 2.5 AI Video Generator

Create Seedance 2.5 videos online on Grok AI. Generate a 5-second clip from a text prompt, a first frame, or a start and end frame, with width and height set by the format you choose.

Model

Seedance 2.5

Clip length

5 seconds

References

30 images · 10 videos · 10 audio

Output

Joint audio and video

Seedance 2.5 Generator

Each clip is 5 seconds. Pick a format, then width and height stay inside the Seedance 2.5 limits.

Describe subject, action, camera, lighting, sound, and how the shot should change over time.

Format & platforms

Choose an aspect ratio. Width stays between 432 and 1684 px. Height stays between 548 and 1632 px.

Duration is fixed at 5 seconds. Width 432–1684. Height 548–1632.

Seedance 2.5 preview appears here after the API is connected.

What is Seedance 2.5?

Seedance 2.5 is ByteDance Seed’s audio-video generation model for one-take storytelling. It builds on the joint audio-video architecture of Seedance 2.0 and focuses on longer clips, multimodal references, and more precise edits.

Each generation on this page is a 5-second clip. Width stays between 432 and 1684 px, and height stays between 548 and 1632 px. Creators can supply up to 30 images, 10 video clips, and 10 audio clips as references, then steer camera language, subject identity, and sound together.

On Grok AI, this page is the Seedance 2.5 workspace for text-to-video, first-frame image-to-video, and start-and-end-frame video. Reference editing, green-screen replacement, and video extension will use the same page once the API is connected.

Seedance 2.5 Capabilities

5-Second Clip

Every Seedance 2.5 generation on this page is 5 seconds. Pick a format and the width and height follow that ratio.

Text to Video

Turn a scene brief into motion, dialogue, ambience, and camera language in one Seedance 2.5 pass.

Image to Video

Lock a first frame, or a start frame and end frame, and let the model travel between them.

Multimodal References

Guide identity, style, motion, and sound with up to 30 images, 10 videos, and 10 audio clips.

Precise Editing

Timestamp-level edits, green-screen replacement, camera changes, and reference-based revisions.

Format Presets

9:16, 16:9, 1:1, 4:5, 3:4, and 21:9. Width stays between 432 and 1684 px, and height stays between 548 and 1632 px.

How to Use Seedance 2.5

  1. 1

    Choose a Seedance 2.5 mode

    Start from text, a first frame, or a start frame plus an end frame.

  2. 2

    Write the shot

    Describe subject, action, camera, lighting, and the sound you want in the clip.

  3. 3

    Choose a format

    Pick an aspect ratio. The clip is 5 seconds, and width and height follow that ratio inside the allowed range.

  4. 4

    Generate and refine

    Run Seedance 2.5, then extend or edit the result once the API connection is live.

Seedance 2.5 vs Seedance 2.0

Seedance 2.5 keeps the audio-video workflow of Seedance 2.0 and expands length, references, and edit control.

CapabilitySeedance 2.5Seedance 2.0
Single clip5-second clips, with width 432–1684 px and height 548–1632 px.Shorter clips, with less room for a full scene in one pass.
ReferencesUp to 30 images, 10 videos, and 10 audio clips, including audio-only references.Fewer image, video, and audio references per request.
EditingTimestamp-level audio and video edits, green screen, camera, and reference edits.Reference-led generation with less precise edit control.
Story continuityStronger shot transitions and scene changes for longer audiovisual stories.Joint audio-video generation aimed at shorter concept clips.

Seedance 2.5 Prompt Examples

Seedance 2.5 FAQ

What is Seedance 2.5?

Seedance 2.5 is ByteDance Seed’s audio-video model. On this page each generation is a 5-second clip, with format presets for width and height.

How long can a Seedance 2.5 video be?

This page generates a 5-second clip. Width is limited to 432–1684 px and height is limited to 548–1632 px.

Does Seedance 2.5 generate audio and video together?

Yes. Seedance 2.5 is a joint audio-video model, so dialogue, ambience, and picture are generated in the same clip.

Can I use images with Seedance 2.5?

Yes. You can start from a first frame or a start frame and end frame. The model also accepts up to 30 reference images, 10 reference videos, and 10 reference audio clips.

Which formats does Seedance 2.5 support?

Choose 9:16, 16:9, 1:1, 4:5, 3:4, or 21:9. Width and height update with the ratio and stay inside 432–1684 by 548–1632. There is no separate resolution setting.

Is Seedance 2.5 available on Grok AI yet?

The Seedance 2.5 page is live. Video generation will start once the API is connected. You can prepare prompts and settings now.

Related AI Video Tools

Text to Video

Generate video from a text prompt with other Grok AI models.

Image to Video

Animate a still image when you already have a key frame.

HappyHorse 1.0

Compare another cinematic text and image video workflow.

MiniMax H3

Start-frame and end-frame video on a different model.

FlashVSR

Upscale a finished clip up to 4x after generation.

Video Prompt Helper

Tighten prompts before you run Seedance 2.5.