Seedance 2.5: FAQ
1. What is Seedance 2.5?
Seedance 2.5 (ByteDance) is the flagship of the Seedance family. It generates video with native synchronized audio and runs in two modes — Keyframe (text-to-video and image-to-video) and Reference (multimodal). Its headline gains over Seedance 2.0 are much longer single-pass clips and a far larger reference pool.
2. What resolution and duration does Seedance 2.5 support?
Resolutions:
480p (faster, cheaper drafts)
720p (balanced)
1080p (highest quality)
No 4K on this model.
Durations (clip lengths):
Any whole number of seconds from 4s to 30s, produced as one continuous generation (longer clips cost more)
3. How does Seedance 2.5 handle audio?
Audio is native, synchronized, and on by default. Seedance 2.5 routes audio by bracket channel, not by quotation marks:
( ) — music and ambient beds
< > — sound effects
{ } — spoken dialogue
【 】 — on-screen subtitles
Declare the speaker, language, and delivery before the braces; the braces hold only the words. Turn audio generation off for a silent clip.
4. Can you combine inputs with Seedance 2.5 in one prompt?
Yes, and this is a major step up. The two modes are mutually exclusive:
Keyframe mode:
A Start Frame with an optional End Frame for interpolation.
Reference mode (multimodal):
Up to 30 reference images
Up to 10 reference videos (each 4–30s; combined ≤30.2s)
Up to 10 reference audio clips (each ~1.8–30.2s; cannot be used alone, must pair with a reference image or video)
50 reference files maximum across all three slots combined
Bind a reference to a spot in the prompt by @-mentioning it; mentions resolve to numbered @Image1 / @Video1 tokens in slot order. Note: you can't mix keyframes and reference inputs in the same generation; pick one mode.
5. What aspect ratios are available?
16:9 (landscape)
9:16 (portrait / vertical)
1:1 (square)
4:3, 3:4, 21:9
auto (the model picks based on your inputs)
6. Are keyframes available with the Seedance 2.5 model?
Yes, in Keyframe mode:
Start Frame: the model animates outward from this image.
End Frame: optional, requires a Start Frame, and drives keyframe interpolation between the two.
Modes at a glance
Mode | What it does |
|---|---|
Keyframe | Text-to-video, or image-to-video with a Start Frame and optional End Frame. |
Reference (multimodal) | Generates from up to 30 images, 10 videos, and 10 audio clips (50 files max). |
Seedance 2.5 vs 2.0
The big jumps: single-pass duration extends to 30s (from 15s), the reference pool grows to 30 images / 10 videos / 10 audio (from 9 / 3 / 3), and audio is directed through bracket channels rather than quotes. Tradeoff: Seedance 2.5 tops out at 1080p compared to 4k with Seedance 2.0.