Muse Video 1.0 Is Live: Text-to-Video and Image-to-Video Parameters, Pricing and Prompts

Muse Video 1.0, the video generation model from Meta Superintelligence Labs, is now available on HiAPI.ai as two models: muse-video-1.0/text-to-video and muse-video-1.0/image-to-video. Every video is 10 seconds at 720p, with sound generated together with the picture.

Pricing

Both models cost the same: $0.0336 per video. Duration and resolution are fixed, so there are no tiers to pick and you know the cost before you submit.

Parameters

Text-to-video: muse-video-1.0/text-to-video

• prompt (required): 1–5000 characters describing the subject, action, camera movement and sound.
• aspect_ratio: 16:9 (default), 9:16, 1:1, 4:3, 3:4, 3:2, 2:3. 16:9 outputs 1280×720; 9:16 outputs 720×1280.

Image-to-video: muse-video-1.0/image-to-video

• prompt (required): 1–5000 characters describing how the subject in the reference images moves, how the camera moves, and the sound you want.
• image_urls (required): 1–9 reference images as http or https URLs. They guide the picture; they are not guaranteed to become a pixel-exact first frame.
• aspect_ratio: auto (default, follows the reference images) or one of the 7 ratios above.

Duration (10 seconds), resolution (720p) and sound are fixed; you don't need to pass them in the request.

Request examples

Text-to-video:

curl https://api.hiapi.ai/v1/tasks \
  -H "Authorization: Bearer $HIAPI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "muse-video-1.0/text-to-video",
    "input": {
      "prompt": "A paper boat drifts across a calm pond at sunrise, low camera following it, gentle ripples, birdsong in the background",
      "aspect_ratio": "16:9"
    }
  }'

Image-to-video (request body):

{
  "model": "muse-video-1.0/image-to-video",
  "input": {
    "image_urls": ["https://example.com/product.jpg"],
    "prompt": "The bottle slowly rotates as soft morning light moves across it, gentle push-in, quiet ambient music",
    "aspect_ratio": "auto"
  }
}

The response contains a task ID. Poll GET /v1/tasks/{taskId} until it completes, then use the video URL.

Prompt tips

1. Start with subject and action: who is in the shot and what they are doing, in one sentence.
2. Then the camera: push-in, tracking, orbit, top-down. Stick to one main camera move per video.
3. Name the sound: videos come with sound, so mention ambience or a music style, such as "birdsong" or "soft piano".
4. Pace it for 10 seconds: one scene with one or two actions works best; don't pack several scenes into one video.
5. Use clear reference images: a well-lit, clearly visible subject keeps its look better; with several images, keep their style consistent.

Common uses

• Vertical 9:16 social clips with ambient sound built in.
• Turn a product photo into a 10-second showcase clip (image-to-video).
• Storyboard and concept previews: try one description in several aspect ratios and pick the framing.

Open the Playground on the model page to try it.