Skip to main content
POST
Muse Video Text to Video
Muse Video is a native-audio video model introduced by Meta Superintelligence Labs. Describe a short scene’s subject, action, camera and sound in a text prompt. These integration-specific model IDs do not claim a direct Meta developer endpoint or a numbered Muse release. 100 FreeModel credits = $1. Both modes have the same per-request price. The completed file determines its actual duration and dimensions. Confirm current availability on the Muse Video model page.
string
required
Use muse-video-t2v.
string
required
A non-empty scene description, at most 100,000 characters under the gateway limit.
integer
default:"10"
Only the JSON number 10 is accepted. Actual output duration is approximately ten seconds.
string
default:"720p"
Only 720p is accepted.
integer
default:"1"
Only 1 is accepted.
boolean
default:"true"
Native sound is always included. Only true is accepted.
string
default:"16:9"
1:1, 16:9, 9:16, 4:3, 3:4, 3:2 or 2:3.
This text mode does not accept image inputs. The route does not expose first/last-frame controls, input video or audio, masks, editing, remix or extension. Unsupported controls are rejected before submission.

Request

Poll and download

Save the returned gateway video id. Poll with the same API key and model ID until status is completed or failed. Download only after completion.
Keep polling an existing task after a client timeout. Repeated status checks do not create a new generation. Save the MP4 promptly and respect any returned expiry time. See the video workflow for task states and error handling.

Open Muse Video

Try text and reference generation with the live per-request price.