MiniMax H3 Video

Tool Overview

MiniMax H3 generates 2K videos from a prompt and optional image, video, and audio references. A request may contain up to 9 images, 3 videos, and 3 audio clips.

Tool Identifier

minimax-h3-video-api

Parameters

Required Parameters

prompt

Type: string
Description: Text description of the video. Refer to uploaded assets as Image 1, Video 1, or Audio 1.

Optional Parameters

duration

Type: number
Default: 5
Supported Values: Any integer from 5 to 15

resolution

Type: string
Default: "2K"
Supported Values: "2K"

aspect_ratio

Type: string
Default: "adaptive"
Supported Values: "adaptive", "21:9", "16:9", "4:3", "1:1", "3:4", "9:16"

reference_image_urls

Type: string[]
Default: []
Limit: Up to 9 public HTTP/HTTPS URLs

reference_video_urls

Type: string[]
Default: []
Limit: Up to 3 public HTTP/HTTPS URLs
Duration: Each video must be 2–15 seconds; combined duration cannot exceed 15 seconds

reference_audio_urls

Type: string[]
Default: []
Limit: Up to 3 public HTTP/HTTPS URLs
Duration: Each audio clip must be 2–15 seconds; combined duration cannot exceed 15 seconds

Audio references may be used without image or video references.

Pricing

2K output costs 150 credits per second.

Request Examples

Prompt Only

curl -X POST https://sjinn.ai/api/un-api/create_tool_task \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "tool_type": "minimax-h3-video-api",
    "input": {
      "prompt": "A cinematic aerial shot above a futuristic coastal city",
      "duration": 5,
      "resolution": "2K",
      "aspect_ratio": "16:9"
    }
  }'

Multimodal References

curl -X POST https://sjinn.ai/api/un-api/create_tool_task \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "tool_type": "minimax-h3-video-api",
    "input": {
      "prompt": "Image 1 is the protagonist. Follow the motion in Video 1 and use Audio 1 as the sound reference.",
      "duration": 10,
      "resolution": "2K",
      "aspect_ratio": "adaptive",
      "reference_image_urls": [
        "https://example.com/character.jpg"
      ],
      "reference_video_urls": [
        "https://example.com/motion.mp4"
      ],
      "reference_audio_urls": [
        "https://example.com/audio.mp3"
      ]
    }
  }'

JavaScript

const response = await fetch('https://sjinn.ai/api/un-api/create_tool_task', {
  method: 'POST',
  headers: {
    Authorization: 'Bearer YOUR_API_KEY',
    'Content-Type': 'application/json',
  },
  body: JSON.stringify({
    tool_type: 'minimax-h3-video-api',
    input: {
      prompt: 'A dancer performs in a room filled with soft volumetric light',
      duration: 8,
      aspect_ratio: '9:16',
      reference_audio_urls: ['https://example.com/music.mp3'],
    },
  }),
});

const result = await response.json();
console.log('Task ID:', result.data.task_id);

Python

import requests

response = requests.post(
    'https://sjinn.ai/api/un-api/create_tool_task',
    headers={
        'Authorization': 'Bearer YOUR_API_KEY',
        'Content-Type': 'application/json',
    },
    json={
        'tool_type': 'minimax-h3-video-api',
        'input': {
            'prompt': 'A dancer performs in a room filled with soft volumetric light',
            'duration': 8,
            'aspect_ratio': '9:16',
            'reference_audio_urls': ['https://example.com/music.mp3'],
        },
    },
)

result = response.json()
print('Task ID:', result['data']['task_id'])

Response

The create endpoint returns a pending task ID. Poll query_tool_task_status until the task completes or fails.

{
  "success": true,
  "errorMsg": "",
  "error_code": 0,
  "data": {
    "task_id": "550e8400-e29b-41d4-a716-446655440000"
  }
}

MiniMax H3 Video