Models / Video generation / minimax-hailuo-2-3

MiniMax Hailuo 2.3

MiniMax Hailuo 2.3 is a general video generation model, making it suitable for stable video creation tasks.

General video

Calling the API

Submit a task to MindVideo with your platform model slug. The response includes a request ID, task ID, normalized status, and estimated cost.

For production retries, send a unique Idempotency-Key per business request. Reusing the same key with the same payload returns the original task without charging again; reusing it with different input returns 409.

Submit a request

curl -X POST https://api.mindvideo.ai/api/v1/tasks \
  -H "Authorization: Bearer mv_live_xxx" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: task_$(date +%s)" \
  -d '{
  "model": "minimax-hailuo-2-3",
  "input": {
    "prompt": "A cinematic portrait shot with smooth camera movement and realistic lighting.",
    "resolution": "768p",
    "duration": 6
  }
}'

Authentication

Every server-side request must include an API key in the Authorization header. Keys can be created, disabled, and deleted from the dashboard.

Keep API keys server-sideDo not expose production API keys in browser code. Route user requests through your own backend before calling MindVideo.

Queue and polling

The platform maps runtime states into queued, processing, succeeded, failed, or cancelled so clients do not depend on model-specific fields.

Submit a request

Create an async generation task and poll the task endpoint for terminal status updates.

Fetch status

curl https://api.mindvideo.ai/api/v1/tasks/task_demo_model \
  -H "Authorization: Bearer mv_live_xxx"

Fetch task events

curl https://api.mindvideo.ai/api/v1/tasks/task_demo_model/events \
  -H "Authorization: Bearer mv_live_xxx"

Read the result

{
  "request_id": "req_demo_model",
  "task_id": "task_demo_model",
  "status": "succeeded",
  "cost_estimate": 0.3294,
  "cost_actual": 0.3294,
  "output": {
    "videos": [
      {
        "url": "https://example.com/output/demo.mp4"
      }
    ]
  }
}

Files

Inputs can reference hosted files or compact data URIs depending on model constraints. Production apps should prefer stable hosted URLs.

Data URI

Useful for small test assets and local playground experiments.

Hosted URL

Recommended for production workloads, retries, and larger image or video inputs.

Schema

Input

{
  "type": "object",
  "required": [
    "prompt"
  ],
  "properties": {
    "prompt": {
      "type": "string",
      "title": "Prompt",
      "minLength": 1,
      "description": "Video content description.",
      "x-playground": {
        "rows": 6,
        "widget": "textarea"
      }
    },
    "duration": {
      "enum": [
        6,
        10
      ],
      "type": "integer",
      "title": "Duration",
      "default": 6
    },
    "watermark": {
      "type": "boolean",
      "title": "Watermark",
      "default": false
    },
    "resolution": {
      "enum": [
        "768p",
        "1080p"
      ],
      "type": "string",
      "title": "Resolution",
      "default": "768p"
    },
    "prompt_optimizer": {
      "type": "boolean",
      "title": "Prompt optimizer",
      "default": false
    },
    "fast_pretreatment": {
      "type": "boolean",
      "title": "Fast pretreatment",
      "default": false,
      "x-playground": {
        "advanced": true
      }
    },
    "first_frame_image": {
      "type": "string",
      "title": "First frame image",
      "description": "Optional first-frame image URL or data URI for image-to-video generation.",
      "x-playground": {
        "accept": "image/*",
        "widget": "file-upload",
        "advanced": true,
        "uploadKind": "image"
      }
    }
  },
  "additionalProperties": false
}

Output example

{
  "request_id": "req_demo_model",
  "task_id": "task_demo_model",
  "status": "succeeded",
  "cost_estimate": 0.3294,
  "cost_actual": 0.3294,
  "output": {
    "videos": [
      {
        "url": "https://example.com/output/demo.mp4"
      }
    ]
  }
}