Image-to-Video

Overview

• API: Image-to-Video
• Description: Use an image as the starting frame and generate a short video based on a text description.
• Model family: Hailuo (MiniMax-Hailuo series)

Authentication

• Use Bearer Token in the request header: Authorization: Bearer <token>.

Request Parameters

• Request Content-Type: application/json

• Request body fields:

  • model (string, required)

    • Model name. Valid values:
      • MiniMax-Hailuo-2.3
      • MiniMax-Hailuo-2.3-Fast
      • MiniMax-Hailuo-02
      • I2V-01-Director
      • I2V-01-live
      • I2V-01

  • prompt (string)

    • Text description for the video, up to 2000 characters. For MiniMax-Hailuo-2.3, MiniMax-Hailuo-2.3-Fast, MiniMax-Hailuo-02, and I2V-01-Director, you can use [directive] syntax to control camera motion.
    • Add camera motion instructions using the [directive] format to achieve precise shot control.

    Supported camera motion directives:

    • Move left/right: [左移], [右移]
    • Pan left/right: [左摇], [右摇]
    • Dolly in/out: [推进], [拉远]
    • Rise/fall: [上升], [下降]
    • Tilt up/down: [上摇], [下摇]
    • Zoom in/out: [变焦推近], [变焦拉远]
    • Other: [晃动], [跟随], [固定]

    Usage rules:

    • Combined motion: multiple directives inside one pair of [] are applied simultaneously, e.g. [左摇,上升]. Avoid combining more than three directives.
    • Sequential motion: directives appear in prompt order, e.g. "...[推进], then...[拉远]".
    • Natural language is also supported, but standard directives are more accurate.

    For tutorials and templates, see the camera motion usage guide.

  • prompt_optimizer (boolean, optional, default true)

    • Whether to auto-optimize the prompt. Set to false for finer control.

  • fast_pretreatment (boolean, optional, default false)

    • Whether to shorten prompt_optimizer runtime. Effective only for MiniMax-Hailuo-2.3, MiniMax-Hailuo-2.3-Fast, and MiniMax-Hailuo-02.

  • duration (integer, required)

    • Video duration in seconds, default is 6. Available values depend on model and resolution:

      Model	720P	768P	1080P
      MiniMax-Hailuo-2.3	-	6 or 10	6
      MiniMax-Hailuo-2.3-Fast	-	6 or 10	6
      MiniMax-Hailuo-02	-	6 or 10	6
      Other models	6	-	6

  • resolution (string, required)

    • Video resolution. Available values depend on model:

      Model	6s	10s
      MiniMax-Hailuo-2.3	768P (default), 1080P	768P (default)
      MiniMax-Hailuo-2.3-Fast	768P (default), 1080P	768P (default)
      MiniMax-Hailuo-02	512P, 768P (default), 1080P	512P, 768P (default)
      Other models	720P (default)	not supported

  • callback_url (string, optional)

    • Callback URL to receive task status updates. When configured, MiniMax will send a POST request to the URL containing a challenge field. The server must return the same challenge value within 3 seconds to complete verification.
    • After verification succeeds, MiniMax will push task status updates to the URL whenever the task state changes. The payload matches the query API response structure.
    • The callback status can be processing, success, or failed.

  • first_frame_image (string, required)

    • Use the specified image as the starting frame of the video. Supports public URLs or Base64 data URLs (data:image/jpeg;base64,...).
    • Image requirements:
      • Formats: JPG, JPEG, PNG, WebP
      • Size: under 20MB
      • Dimensions: short side greater than 300px, aspect ratio between 2:5 and 5:2

Request Example

curl -X POST "https://api.gpt.ge/task/minimax/v1/video_generation" \
  -H "Authorization: Bearer $YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "MiniMax-Hailuo-02",
    "prompt": "A girl climbs onto the horse [右移], then pulls the horse's head and turns, galloping away [拉远].",
    "duration": 10,
    "resolution": "768P",
    "first_frame_image": "https://p2.a.kwimgs.com/bs2/upload-ylab-stunt/ai_portal/1731125871/6N0QrAnAeU/409-f054fa25c21a.png"
  }'

Request Body Example

{
  "model": "MiniMax-Hailuo-02",
  "prompt": "A girl climbs onto the horse [右移], then pulls the horse's head and turns, galloping away [拉远].",
  "duration": 10,
  "resolution": "768P",
  "first_frame_image": "https://p2.a.kwimgs.com/bs2/upload-ylab-stunt/ai_portal/1731125871/6N0QrAnAeU/409-f054fa25c21a.png"
}

Success Response Example

{
  "task_id": "106916112212032",
  "base_resp": {
    "status_code": 0,
    "status_msg": "success"
  }
}

Response Fields

• task_id (string): task ID to query status and results.
• base_resp (object): basic response information
  • status_code (integer): 0 indicates success.
  • status_msg (string): status message.

Notes

• Make sure required fields are included. Missing required parameters may prevent the model name from being recognized.
• For models that support directive syntax, using standard directives in the prompt improves camera motion control.
• See the request parameter section for callback and image requirements.