{"info":{"title":"bytedance/latentsync","version":"1.0.0"},"paths":{"/":{"post":{"summary":"Generate","responses":{"200":{"content":{"application/json":{"schema":{"$ref":"#/components/schemas/Output"}}},"description":"Successful Response"}},"operationId":"bytedance_latentsync_generate","requestBody":{"content":{"application/json":{"schema":{"$ref":"#/components/schemas/Input"}}},"required":true}}}},"openapi":"3.1.0","components":{"schemas":{"Input":{"type":"object","title":"Input","required":["video_url","audio_url"],"properties":{"seed":{"type":"integer","title":"Seed","default":0,"x-order":3,"description":"Random seed. 0 (the default) draws a fresh random seed on every run, so repeated calls with identical inputs differ; any positive integer is used as given for a repeatable run."},"audio_url":{"type":"string","title":"Audio URL","format":"uri","x-order":1,"description":"URL of the speech track the speaker should appear to say (MP3, AAC, WAV or M4A). The result runs for the shorter of this track and the video, trimmed down to a whole multiple of 0.64 seconds — supply audio slightly shorter than the video and expect the last fraction of a second to be cut."},"video_url":{"type":"string","title":"Video URL","format":"uri","x-order":0,"description":"URL of the source video — an MP4 with one clearly visible, front-facing speaker in every frame; the run fails on any frame where no face is detected. The result keeps this video's frame size, and its identity, lighting and background; only the mouth region is regenerated. Video running past the end of the audio is discarded, so trim it to roughly the audio's length."},"guidance_scale":{"type":"number","title":"Guidance Scale","default":1,"maximum":10,"minimum":0,"x-order":2,"description":"Strength of the audio conditioning during diffusion. The default of 1 leaves classifier-free guidance off; values above 1 switch it on. The model's own demo exposes 1–3.5, though the field accepts up to 10."}}},"Output":{"type":"string","title":"Output","format":"uri","description":"URL of the generated MP4: the source footage with the speaker's mouth re-animated to the supplied audio, carrying that audio as its soundtrack, at the input's frame size and 25 fps."}}}}