Seedance 2.5 - Coming Soon!
Seedance 2.5, ByteDance's newest and most powerful video model, generates up to 30 seconds of video from 50 multimodal references — with synchronized audio, all in a single pass.
Frequently asked questions
Seedance 2.5 by ByteDance is an AI video generator that creates native 30-second video clips using text, images, video, and audio references. It generates synchronized audio in the same pass as the visuals and is available inside Artlist's AI Toolkit.
Seedance 2.5 generates clips up to 30 seconds (vs. 15s), accepts 50 multimodal references (vs. 12), adds audio-only references, adaptive aspect ratio, and Smart Duration. Resolution starts at 720p — Seedance 2.0 offers up to 4K for resolution-priority workflows.
Each clip is 4 to 30 seconds. Set the exact duration manually or use Smart Duration to let the model auto-select a natural length in the 4–15 second range. Duration affects credit usage.
Yes. Audio — dialogue, music, and sound effects — is co-generated in the same latent space as the visuals. Sound stays in sync because it was never generated separately. No post-alignment or second pass needed.
Up to 30 images, 10 video clips, and 10 audio files per generation. Audio-only references are new in 2.5. Assign each reference by role — subject, style, camera, motion, dialogue, music, or SFX — through the prompt.
480p and 720p at 24 FPS at launch. 1080p output is expected to follow. For 4K output today, use Seedance 2.0 or Seedance 4K inside the Artlist AI Toolkit.
Still have questions? We're here to help.