Directed motion
Describe camera movement, pacing and action in natural language for shots with a clear visual intention.
The fast tier of ByteDance’s video model — same fields as Seedance 2.5 at 480p/720p, with flat per-second pricing for high-volume generation and quick iteration.
A flexible video creation model designed for motion, composition and cinematic direction. Start with a sentence, a frame, or a small set of references and shape the shot you want.
From a first visual idea to a finished clip, the workflow stays expressive and practical.
Describe camera movement, pacing and action in natural language for shots with a clear visual intention.
Use first and last frames to guide transitions, poses and scene changes with stronger visual control.
Bring in images, video and audio references when you need a consistent look, subject or atmosphere.
Generate up to 30 seconds in one request for narrative beats, explainers and ad concepts.
Choose landscape, portrait, square, ultrawide or adaptive framing for the channel you publish on.
Prompt, configure, estimate cost and generate from one focused workspace with no extra setup.
Four steps from idea to downloadable result.
Explain the subject, action, camera and mood. Specific verbs and visual details lead to better results.
Upload a first frame, last frame or reference media. Choose one workflow based on the kind of control you need.
Select resolution, aspect ratio, duration, audio and output format. The studio updates your estimated cost.
Start generation, review the result in the preview panel and iterate with a clearer prompt when needed.
Frame workflows cannot be combined with reference media. Reference videos are limited to 30 seconds total.
Use MP4 for broad compatibility or MOV when your downstream workflow calls for it.
Answers to common questions about inputs, settings and pricing.
Seedance 2.5 Flash is the fast tier of ByteDance’s Seedance 2.5 video model family, released November 12, 2026. It uses the same fields as Seedance 2.5 and outputs 480p or 720p — built for high-volume generation and quick iteration.
Choose 480p or 720p and generate clips from 4 to 30 seconds. For 1080p output, generate a new version with Seedance 2.5 from the same prompt and references.
A flat per-second rate regardless of audio: 30 gems per second at 480p and 60 gems per second at 720p. Without video input, cost is the rate multiplied by output duration. With reference video input, cost is the rate multiplied by input duration plus output duration.
Yes. You can attach up to 5 reference videos of 2 to 30 seconds each, with a combined duration of up to 30 seconds. Reference video duration is included in billing.
Yes — a Flash-exclusive. Attach an audio reference with no image or video, describe the scene, and it generates visuals that match the music.
Up to 10 reference images, 5 reference videos and 5 audio clips per request. Seedance 2.5 takes more (30/10/10) if you need heavier reference sets.
No. The first/last-frame workflow and reference-media workflow are mutually exclusive, following the provider API rules.
Start with a prompt or upload a frame and make a motion-rich clip in the Seedance 2.5 Flash Studio.
Start creating with Seedance 2.5 Flash