seedance-2.0-fast text-to-video capability illustrated
A short film with a live-action feel demonstrates the text-to-video path of seedance-2.0-fast: from a text prompt alone, it generates in a single pass a finished clip with camera choreography, physical behavior, and native audio. The work's inputs map strictly to the fields of the author's pipeline—prompt describes the scene and camera movement, duration, resolution, and aspect_ratio set the specifications, generate_audio controls native audio, and web_search stays off to verify pure generation capability; the first-frame, last-frame, and reference-material fields are not used, so this entry showcases the lowest-cost, text-only-driven path. It suits rapid validation of creative prototypes and storyboards, batch variants of short social media videos, and low-cost trial and error before formal production; when a character or product texture needs to be locked in, the reference-material fields of the same pipeline can be used instead.
Style Text-to-Video · Product Demo · Cinematic
Model seedance-2.0-fast
Try-it is unavailable: this model's playground configuration isn't ready yet.
Prompt
A single continuous cinematic shot showing text-generated video: a sleek matte-black concept car glides slowly down a wet, neon-lit city street at night, rain streaking across the frame and reflections stretching over the asphalt. The camera starts from a low angle close to the ground, tracks along the side of the car, then rises smoothly to reveal the empty avenue and glowing signs in a wide 16:9 frame. Shallow depth of field, anamorphic lens flares, cool blue and magenta lighting, fine raindrop particles, realistic motion blur, and mist kicked up by the tires. A steady, unhurried pace held for five seconds; no cuts; accompanied by a subtle low hum of city ambience.
Copy the full prompt and adapt it for your own generation.
Tags