Seedance 2.5

bytedance/seedance-2.5

One model id that selects text, image, multimodal-reference, video-edit, or video-extend generation from the supplied inputs.

Pricing

Includes our margin. Charged per request from your balance.

Video reference billing: input plus output seconds; each reference video is billed for at least 2 s, 30 s combined

Parameters this model accepts.

Parameter	Values	Default
`mode` Automatically infer the operation from inputs, or explicitly select video extension.	Automatic (from inputs), Text / references to video, Image to video, Video edit, Video extend Default `auto`	`auto`
`duration` Output duration from 4 to 30 seconds	4–30 seconds Default `5`	`5`
`resolution` Output resolution	480p, 720p, 1080p, 4K Default `720p`	`720p`
`aspect_ratio` Used for text and multimodal-reference generation; image and video inputs determine framing.	Ultra-wide (21:9), Landscape (16:9), Classic (4:3), Square (1:1), Portrait classic (3:4), Portrait (9:16) Default `16:9`	`16:9`
`generate_audio` Generate synchronized sound effects, ambience, and speech.	On or off Default `on`	`on`