MiniMax H3
hailuo-h3
Combine text, images, video and audio references into 768p or 2K video. Guide generation with first and last frames, or up to 9 images, 3 videos and 3 audio references.
- Output resolution
- 768P / 2K
- Output duration
- 5 to 15 seconds
- Reference inputs
- 9 images / 3 videos / 3 audio files
Playground
Tune parameters and generate a video right on this page. Fill in your request first and sign in when you submit. Your parameters will be kept.
Output
All tasks ↗Your next scene starts here
Choose a model, write a prompt and generate. Your video will appear here when ready.
Model capabilities
Text to video
Start with a prompt. Describe your subject, action, scene and camera movement without preparing any source media.
First and last frames
Set the opening with a first frame and optionally guide the ending with a last frame. Frames cannot be mixed with reference lists.
Reference to video
Combine image, video and audio URLs. Describe the role of each reference in your prompt.
Build with the API
Tune your parameters in Playground, then copy the request code into your application.
curl -X POST 'https://api.xbaiapi.com/api/v1/jobs/createTask' \
-H 'Authorization: Bearer YOUR_API_KEY' \
-H 'Content-Type: application/json' \
-d '{
"model": "hailuo-h3",
"input": {
"prompt": "A lighthouse by the sea at sunset. The camera slowly moves forward.",
"duration": 5,
"resolution": "2k",
"ratio": "16:9"
}
}'From a first experiment to your API
Run real tasks using your account balance, track generation and preview the resulting video.
Results are retained for 30 days. Failed tasks refund the reserved credits automatically. Track every task in your console.