ByteDance's next-generation motion AI. Create up to 30-second videos from text, images and multimodal references — with native audio and adaptive framing.
Select the frame dimensions. Default is 16:9.
Standard (480p) / High (720p)
Seedance 2.5 supports flexible durations from 4s up to 30s.
Ready to Create Your Seedance 2.5 Video
Seedance 2.5 is ByteDance's next-generation AI video generation model, built for longer, richer and more controllable video creation. It generates expressive clips of up to 30 seconds from plain text, a first frame image, or a full set of multimodal references — images, videos and audio working together.
Compared with Seedance 2.0, version 2.5 doubles the maximum duration, understands reference media far more accurately, supports adaptive framing and native audio generation, and delivers sharper motion with better physics — everything you need for production-ready AI video workflows.
Tell complete stories in a single generation. Seedance 2.5 natively supports clips from 4 to 30 seconds — double the range of Seedance 2.0 — with smoother story flow and richer scenes.
Attach reference images, videos and audio to keep characters, products, colors and styles consistent across every scene. Cite them directly in your prompt with @Image1, @Video1 or @Audio1.
Define exactly where your video starts and ends. Upload a first frame and an optional last frame, and Seedance 2.5 choreographs the motion between them with frame-accurate continuity.
Choose from 16:9, 9:16, 1:1, 4:3, 3:4, 21:9 or Adaptive mode that detects the best framing from your input media — perfect for YouTube, TikTok, Reels and cinematic formats alike.
Generate synchronized sound effects and ambience together with the picture. Toggle audio on for immersive results, or off to save credits on silent drafts.
Seedance 2.5 delivers next-generation motion quality — more realistic physics, cleaner details and improved prompt adherence, ready for professional and commercial projects.
Start from a pure text prompt, or switch to Image to Video to upload a first frame, an optional last frame, and reference images, videos or audio for guided generation.
Write a detailed prompt — type @ to cite your references — then pick an aspect ratio, 480p or 720p resolution, a duration up to 30 seconds, MP4 or MOV output, and whether to generate audio.
Click Generate and let Seedance 2.5 render your video. Preview the result in the page, download the file, or revisit any past creation from your History tab.
Seedance 2.5 is ByteDance's next-generation AI video generation model. It creates videos of up to 30 seconds from text prompts, first/last frame images, and multimodal references (images, videos, audio), with optional native audio generation and adaptive aspect ratios.
Seedance 2.5 doubles the maximum duration to 30 seconds, adds more accurate multimodal reference understanding, supports adaptive aspect ratios and MP4/MOV output, and delivers sharper motion with better physics compared with Seedance 2.0.
Seedance 2.5 supports 480p and 720p output with flexible durations from 4 to 30 seconds, and aspect ratios of 16:9, 9:16, 1:1, 4:3, 3:4, 21:9 or Adaptive.
Pricing is per second of output: 480p costs 40 gems/s (50 with audio) and 720p costs 80 gems/s (120 with audio). When you attach reference videos, the input video duration is added to the billed duration.
Yes. Seedance 2.5 can generate synchronized sound effects and ambience natively. You can toggle audio generation off at any time to reduce the per-second cost.
Upload reference media in Image to Video mode and cite them in your prompt with @Image1, @Video1 or @Audio1. Seedance 2.5 uses them to keep characters, products, styles and sound consistent throughout the generated video.
