ThinkSound mark

ThinkSound Vid2Vid

Generate realistic, synchronized audio for a video clip, optionally guided by a text prompt. This model is useful for adding sound effects and ambient soundbeds that match on-screen action, and pricing is based on inference time (roughly proportional to video length and compute).

Inputs

Video

Outputs

Audio

Lab

fal-ai

Launch your first agent today

Build without code. Share without limits.