seedance-2.5/text-to-video Token

Seedance 2.5 Standard text-to-video: generate high-quality video from a text prompt with native 480p/720p and upscaled 1080p/2k/4k, native audio, flexible ratio and duration. Token-based billing.

Playground

Input

*API Key

Please select API Key

*prompt

A cinematic sequence on a rainy night in a southern Chinese old town: street-life smoke and neon, shallow depth of field, Hong Kong-film warm/cool contrast.  [00:00-00:07] Just after rain at an old-town gate, red lanterns recede into the street, wet ground smearing warm sign reflections; ground-level slow push as pedestrians cross, soft-focus silhouettes in the foreground. Ambience: eave drips, crowd, distant scooter. [00:07-00:14] Close-up: a middle-aged man in a dark jacket lights a cigarette at a food stall, flame briefly revealing his profile and crow's feet, steam rising from bamboo steamers; shallow DOF neon bokeh. Ambience: ladle, oil sizzle, lighter. [00:14-00:20] He looks toward the street end; camera pans with his gaze to the river, opposite skyline lights shattered on the water. [00:20-00:25] Slow rise to a high half-angle; lanterns, steam and crowd fuse into warm haze; fade.  Style: cinematic, 35mm shallow DOF, warm neon vs cool night, strong wet-ground speculars, gentle grain, realistic crowds and skin. Camera: ground push, shallow close-up, gaze pan, slow rise — all stable. Audio: ambience first (rain, voices, stalls, scooter), low electronic bed. No dialogue. Avoid: readable brand logos or legible sign text; subtitles, watermarks; deformed faces or hands; duplicating/clipping pedestrians; jump cuts or flicker.

*resolution

720p

*duration

generateAudio

watermark

ratio

Adaptive

webSearch

returnLastFrame

bitrateMode

standard

seed

outputFormat

mp4

draft

seedance-2.5/text-to-video Token

Input

Result