# MiniMax H3 (Hailuo AI) - AI Video Workflows, Prompts & Local Deployment Hub

> **Official Hub for MiniMax H3 (`minimaxh3.run`)** — An all-in-one open-source resource platform for **MiniMax H3 (Hailuo AI 3.0)**, the state-of-the-art native multi-modal 2K AI video generation model with synchronized 3D stereo audio.

[![License: MIT](https://img.shields.io/badge/License-MIT-blue.svg)](LICENSE)
[![ComfyUI Supported](https://img.shields.io/badge/ComfyUI-Supported-green.svg)](https://github.com/comfyanonymous/ComfyUI)
[![Hugging Face](https://img.shields.io/badge/%F0%9F%A4%97-Hugging%20Face-orange)](https://huggingface.co)
[![MiniMax H3 Model](https://img.shields.io/badge/Model-MiniMax--H3-purple)](#)

---

## 🚀 Key Features

- 🎨 **Interactive ComfyUI Workflow Builder**: Export production-ready `.json` workflows for Text-to-Video (T2V), Image-to-Video (I2V), Reference-to-Video (Ref2VA), and Native Stereo Audio Synchronization.
- ✍️ **Prompt Engineering Studio**: Formula-based prompt builder tailored for MiniMax H3 native audio triggers, camera movements, lighting, and 2K resolution tags.
- ⚡ **GPU VRAM & Hardware Calculator**: Instant compatibility checker for RTX 3060 (12GB), RTX 4080 (16GB), RTX 4090 (24GB), A100, and Apple Silicon with GGUF / FP8 / INT8 quantization support.
- 💻 **API Playground & Code Generator**: Ready-to-use Python, TypeScript, cURL, and Fal.ai code snippets for MiniMax H3 API integration.
- 📊 **Model Comparison Matrix**: Benchmark comparisons between MiniMax H3, Seedance 2.5, Wan 2.1, Kling AI, Sora, and CogVideoX.
- 🌐 **Chinese AI Video Generator Directory**: Complete guide for international creators to access Chinese AI models without native phone registration, plus Faceless Channel guides.

---

## 📦 Quick Start

### Prerequisites
- Node.js 18.x or higher
- npm or pnpm

### Installation

```bash
# Clone the repository
git clone https://github.com/your-username/minimax-h3-hub.git

# Navigate into the project directory
cd minimax-h3-hub

# Install dependencies
npm install

# Start the development server
npm run dev
```

Visit `http://localhost:3000` to access the hub locally.

---

## 🧩 ComfyUI Setup for MiniMax H3

1. Install [ComfyUI](https://github.com/comfyanonymous/ComfyUI).
2. Install custom nodes via ComfyUI Manager or git clone:
   ```bash
   cd ComfyUI/custom_nodes
   git clone https://github.com/Comfy-Org/ComfyUI-MiniMaxH3.git
   ```
3. Download weights from Hugging Face or ModelScope into `ComfyUI/models/checkpoints/`:
   - `minimax_h3_video_vae_fp16.safetensors`
   - `minimax_h3_audio_vae_fp32.safetensors`
   - `qwen3vl_32b_minimax_h3.safetensors` (Text Encoder)
4. Launch ComfyUI and import the generated `.json` workflow from `minimaxh3.run`.

---

## 💻 Tech Stack

- **Framework**: React 19 + TypeScript
- **Styling**: Tailwind CSS v4
- **Icons**: Lucide React
- **Animations**: Motion (Framer Motion)
- **Build Tool**: Vite

---

## ⚠️ Disclaimer

`minimaxh3.run` is an independent, open-source community resource for AI video workflows and guides. We are not affiliated with, sponsored by, or endorsed by MiniMax Inc. or Hailuo AI. All product names, logos, and trademarks belong to their respective owners.

---

## 📄 License

This project is released under the [MIT License](LICENSE).