MiniMax H3 多模态(图 + 视频 + 音频)

开发中

POST

/v1/video/generations

MiniMax H3（Hailuo 03）视频生成模型，支持文本/图片/视频/音频多模态输入
视频生成为异步任务：先 POST 创建任务拿到 task_id，再轮询 GET /v1/videos/{task_id} 查询结果；成功后从响应中取视频地址

请求参数

Header 参数

Body 参数application/json必填

示例

{
  "model": "MiniMax-H3",
  "prompt": "参考人物在雪山前讲话，保持其声音与形象",
  "images": ["https://example.com/person.jpg"],
  "videos": ["https://example.com/motion.mp4"],
  "audios": ["https://example.com/voice.wav"],
  "duration": 8,
  "metadata": {
    "resolution": "2K",
    "ratio": "adaptive"
  }
}

请求示例代码

Shell

JavaScript

Java

Swift

PHP

Python

HTTP

Objective-C

Ruby

OCaml

Dart

curl --location '/v1/video/generations' \
--header 'Accept: application/json' \
--header 'Authorization: Bearer {{YOUR_API_KEY}}' \
--header 'New-Api-User: 1' \
--header 'Content-Type: application/json' \
--data '{
  "model": "MiniMax-H3",
  "prompt": "参考人物在雪山前讲话，保持其声音与形象",
  "images": ["https://example.com/person.jpg"],
  "videos": ["https://example.com/motion.mp4"],
  "audios": ["https://example.com/voice.wav"],
  "duration": 8,
  "metadata": {
    "resolution": "2K",
    "ratio": "adaptive"
  }
}'

返回响应

🟢200

application/json

任务提交成功

Bodyapplication/json

示例

{
    "id": "task_BEeh4v26c0x9JFuNCtl7QHTcjRTEwfEB",
    "task_id": "task_BEeh4v26c0x9JFuNCtl7QHTcjRTEwfEB",
    "object": "video",
    "model": "MiniMax-H3",
    "status": "queued",
    "progress": 0,
    "created_at": 1787121172
}

修改于 2026-09-10 08:30:04

MiniMax H3 图生视频(首尾帧)

查询视频生成任务状态