# Hailuo Video

> Hailuo Video is an Agent Skill for documents published in the majiayu000/claude-skill-registry repository on GitHub: a SKILL.md instruction file an AI agent can load to do this task.

Generate AI videos with Hailuo (MiniMax) via AceDataCloud API. Use when creating videos from text descriptions or animating images into video. Supports text-to-video and image-to-video with director mode for precise control.

- Skill name: `hailuo-video`
- Category: [Documents](https://www.openparable.com/skill/documents/1)
- Source: https://github.com/majiayu000/claude-skill-registry/blob/HEAD/skills/ai-ml/hailuo-video-acedatacloud-skills-2/SKILL.md
- Repository: majiayu000/claude-skill-registry (498 GitHub stars when collected)

## SKILL.md

# Hailuo Video Generation

Generate AI videos through AceDataCloud's Hailuo (MiniMax) API.

## Authentication

```bash
export ACEDATACLOUD_API_TOKEN="your-token-here"
```

## Quick Start

```bash
curl -X POST https://api.acedata.cloud/hailuo/videos \
  -H "Authorization: Bearer $ACEDATACLOUD_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"action": "generate", "prompt": "a dolphin jumping through ocean waves at golden hour", "model": "minimax-t2v"}'
```

## Models

| Model | Type | Best For |
|-------|------|----------|
| `minimax-t2v` | Text-to-Video | Creating video from text description |
| `minimax-i2v` | Image-to-Video | Animating a still image |
| `minimax-i2v-director` | Image-to-Video (Director) | Precise control over animation from image |

## Workflows

### 1. Text-to-Video

```json
POST /hailuo/videos
{
  "action": "generate",
  "prompt": "a time-lapse of flowers blooming in a meadow",
  "model": "minimax-t2v"
}
```

### 2. Image-to-Video

Animate a still image into a video clip.

```json
POST /hailuo/videos
{
  "action": "generate",
  "prompt": "gentle wind blows through the scene",
  "model": "minimax-i2v",
  "first_image_url": "https://example.com/landscape.jpg"
}
```

### 3. Image-to-Video (Director Mode)

More precise control over the animation.

```json
POST /hailuo/videos
{
  "action": "generate",
  "prompt": "camera slowly zooms in while leaves fall gently",
  "model": "minimax-i2v-director",
  "first_image_url": "https://example.com/scene.jpg"
}
```

## Parameters

| Parameter | Required | Values | Description |
|-----------|----------|--------|-------------|
| `action` | Yes | `"generate"` | Action type |
| `prompt` | Yes | string | Video description |
| `model` | Yes | `"minimax-t2v"`, `"minimax-i2v"`, `"minimax-i2v-director"` | Model |
| `first_image_url` | For i2v | string | Source image URL (required for image-to-video) |
| `mirror` | No | boolean | Mirror the output |
| `callback_url` | No | string | Async callback URL |

## Task Polling

```json
POST /hailuo/tasks
{"task_id": "your-task-id"}
```

States: `processing` → `succeed` or `failed`.

## Gotchas

- `first_image_url` is **required** for `minimax-i2v` and `minimax-i2v-director` models
- Director mode (`minimax-i2v-director`) provides finer camera/motion control than standard i2v
- The `action` field currently only supports `"generate"` — no extend or edit
- Flat pricing per generation regardless of model
- Use `mirror: true` to horizontally flip the output if needed

---

Source: https://www.openparable.com/skill/documents/hailuo-video (OpenParable). Cite the canonical URL above.