Logo
Image
Video
Voice
Apps
Motion Control
Avatar StudioNew
Marketing StudioNew
More
New
Other
Pricing

Turn your idea into video

Ready to create some AI magic? Describe your scene, pick your settings, and watch as AI brings your vision to life in stunning video.

Write a detailed prompt describing your video scene
Optionally upload reference images or videos
Hit generate and create cinematic videos in minutes
Start creating on the left
fal.ai

Native 30-second cinematic video with audio

Seedance 2.5 is ByteDance's longest-form video model on Zyka. Generate a single continuous shot of up to 30 seconds at 480p or 720p, with native audio, up to 30 reference images, 10 reference videos, and 10 reference audio tracks.

  • Up to 30 seconds in a single pass
  • Native audio generation
  • 30 image / 10 video / 10 audio references
  • 480p and 720p output

How Seedance 2.5 Works

01

Write a longer scene

Seedance 2.5 renders up to 30 seconds in one shot, so the prompt can carry a full beat — setup, movement, and resolution — instead of a single moment.

02

Attach references

Upload up to 30 images, 10 videos, and 10 audio tracks, then call them in the prompt as @Image1, @Video1, @Audio1 to steer identity, motion, and sound.

03

Pick length and resolution

Choose any duration from 4 to 30 seconds at 480p or 720p. Native audio is generated alongside the visual track in the same pass.

About Seedance 2.5

Seedance 2.5 is ByteDance's long-form video model on Zyka, available via fal.ai. Where the Seedance 2.0 tiers cap out at 15 seconds, 2.5 generates up to 30 seconds natively in a single pass — no stitching, no continuation calls, and no drift between segments.

It also carries the widest reference budget in the Seedance family: 30 reference images, 10 reference videos, and 10 reference audio tracks, all addressable from the prompt with @-mentions. That makes it a strong fit for scenes that need consistent characters, props, or motion style across a longer runtime.

Output is capped at 480p and 720p. When you need 1080p or 4K, use standard Seedance 2.0 — 2.5 trades peak resolution for length and reference depth.

Frequently Asked Questions

What is Seedance 2.5?

Seedance 2.5 is ByteDance's long-form video model — up to 30 seconds of continuous cinematic video with native audio, generated in a single pass from text, images, video, or audio references.

How long can a Seedance 2.5 video be?

Up to 30 seconds, selectable in one-second steps from 4 seconds upward. It is generated as one continuous shot rather than stitched segments.

Does Seedance 2.5 support 1080p or 4K?

No. Seedance 2.5 outputs 480p or 720p. For 1080p or 4K, use standard Seedance 2.0.

How many references can I attach?

Up to 30 images, 10 videos, and 10 audio files. Mention them in the prompt as @Image1, @Video1, and @Audio1 to control which reference drives which part of the scene.

Should I use Seedance 2.5 or 2.0?

Use 2.5 when runtime and reference depth matter — longer scenes, consistent characters, richer audio direction. Use 2.0 when you need maximum resolution, including 1080p and 4K.