Create asynchronous Muse Video tasks from a prompt or one source image.
Muse Video API SDKs for JavaScript, Python, Ruby, Go, and Java on RunAPI.
import { MuseVideoClient } from "@runapi.ai/muse-video";
const client = new MuseVideoClient();
const task = await client.textToVideo.create({
model: "muse-video-v1",
prompt: "A city at dawn",
});
const completed = await client.imageToVideo.run({
model: "muse-video-v1",
prompt: "Animate this scene",
image_url: "https://images.unsplash.com/photo-1470770841072-f978cf4d019e",
});Use create() and get() for asynchronous workflows, or run() to poll until completion. Generated file URLs are temporary; store needed results in durable storage during their retention window.
See the Muse Video catalog for current model details and pricing.