🎨 ConCreat

A cutting-edge web application for creating and managing multimedia content with integrated AI-powered capabilities

📖 Documentation • 🐛 Report Bug • ✨ Request Feature

📋 Table of Contents

✨ Features
🤖 ComfyUI Integration
⚡ Quick Start
🔧 Manual Setup
📂 Project Architecture
🔌 API Reference
🔐 Environment Configuration
🛠️ Development Workflow
🤝 Contributing
📄 License
🛠️ Technology Stack
🙏 Acknowledgments

✨ Features

🎯 Core Capabilities

Feature	Description
💬 AI Chat Assistant	Interactive conversations powered by Ollama local models
🎨 AI Image Generation	Create stunning images using ComfyUI workflows with GGUF models
🎬 Video Processing	Advanced video generation and processing with HunyuanVideo integration
🗣️ Text-to-Speech	Natural voice synthesis powered by Chatterbox technology
🎵 Voice Cloning	Personalize audio content with advanced voice replication
📱 Modern UI/UX	Sleek interface built with Next.js, React, and Tailwind CSS
🔗 ComfyUI Integration	Node-based AI workflows for professional content creation

🚀 Advanced Features

GGUF Model Support - Optimized quantized models for efficient inference
Customizable Pipelines - Node-based workflows that can be modified and extended
High-Quality Output - Support for various formats with configurable quality settings
Prompt Engineering - Advanced text encoding with positive/negative prompts
Real-time Processing - Fast generation with optimized model architectures
Cross-platform Compatibility - Works on Windows, macOS, and Linux

🤖 ComfyUI Integration

ConCreat leverages ComfyUI, a powerful node-based interface for AI image and video generation, to provide advanced creative tools.

📋 Included Workflows

🎨 Image Generation Workflow (`workflows/imagemaker.json`)

Advanced image creation using GGUF models like z_image_turbo, with support for LoRA models and custom prompts

🎬 Video Generation Workflow (`workflows/video.json`)

Video creation using HunyuanVideo15 models for high-quality video generation from images

🔧 Workflow Features

⚡ GGUF Model Support: Optimized quantized models for efficient inference
🔄 Customizable Pipelines: Node-based workflows that can be modified and extended
🎯 High-Quality Output: Support for various image and video formats with configurable quality settings
💬 Prompt Engineering: Advanced text encoding with positive and negative prompts

📥 ComfyUI Model Requirements

To use the included ComfyUI workflows, you'll need to download the following models and place them in your ComfyUI models directory:

🎨 Required Models for Image Generation

Model	Filename	Download Link	Location
VAE	`ae.safetensors`	Hugging Face	`ComfyUI/models/vae/`
CLIP	`Qwen3-4B-UD-Q6_K_XL.gguf`	Hugging Face	`ComfyUI/models/clip/`
Unet	`z_image_turbo-Q8_0.gguf`	Hugging Face	`ComfyUI/models/unet/`

🎬 Required Models for Video Generation

Model	Filename	Download Link	Location
Checkpoint	`HV15-Rapid-AIO-v1.safetensors`	Hugging Face	`ComfyUI/models/checkpoints/`
CLIP Vision	`sigclip_vision_patch14_384.safetensors`	Hugging Face	`ComfyUI/models/clip_vision/`

🛠️ ComfyUI Installation

# Clone ComfyUI repository
git clone https://github.com/comfyanonymous/ComfyUI.git
cd ComfyUI

# Install dependencies
pip install -r requirements.txt

# Download required custom nodes
# ComfyUI-GGUF: https://github.com/city96/ComfyUI-GGUF
# rgthree-comfy: https://github.com/rgthree/rgthree-comfy

💡 Note: Model file sizes can be large (several GB). Ensure you have sufficient disk space and a stable internet connection for downloads.

⚡ Quick Start

📋 System Requirements

Node.js: Version 18 or higher
Python: Version 3.11 or above
Git: For version control
Ollama: Local AI model server (latest version recommended)
Storage: At least 10GB free space for models

🚀 Automated Installation

Clone the Repository

git clone https://github.com/kliewerdaniel/concreat.git
cd ConCreat

Execute Setup Script
```
npm run setup
```
This command handles all dependency installations and creates the Python virtual environment automatically.
Launch Development Server
```
npm run dev
```
Access Application Open your browser and navigate to http://localhost:3000

🔧 Manual Setup

For those preferring step-by-step installation:

📦 Frontend Dependencies

npm install

🐍 Python Environment Setup

python3 -m venv venv
source venv/bin/activate  # Use `venv\Scripts\activate` on Windows

📚 Python Dependencies

pip install -r requirements.txt

🤖 Ollama Setup

ConCreat uses Ollama for local AI chat functionality. Install and set up Ollama:

# Install Ollama (follow instructions for your OS at https://ollama.ai)
# For macOS/Linux:
curl -fsSL https://ollama.ai/install.sh | sh

# For Windows: Download from https://ollama.ai/download

# Pull recommended models
ollama pull gemma      # Default chat model
ollama pull llama2     # Alternative model
ollama pull mistral    # Additional model option

# Start Ollama service (runs in background)
ollama serve

💡 Note: Ollama runs on localhost:11434 by default. The application will automatically detect if Ollama is running and fall back to mock responses if not available.

▶️ Application Launch

npm run dev

📂 Project Architecture

ConCreat/
├── 📁 src/app/
│   ├── 📁 api/
│   │   ├── 📁 chat/          # 💬 Chat system endpoints
│   │   ├── 📁 generate/      # 🎨 Content creation APIs
│   │   ├── 📁 images/        # 🖼️ Image manipulation APIs
│   │   ├── 📁 tts/           # 🗣️ Text-to-speech conversion
│   │   ├── 📁 videos/        # 🎬 Video processing endpoints
│   │   └── 📁 voices/        # 🎵 Voice management system
│   ├── 🎨 globals.css        # Global stylesheet
│   ├── 📱 layout.tsx         # Application layout component
│   └── 🏠 page.tsx           # Main page component
├── 🌐 public/                # Static resources
├── 🔧 workflows/             # Workflow configuration files
├── 🐍 tts_service.py         # Python TTS service implementation
├── ⚙️ setup.sh               # Automated setup script
├── 📋 requirements.txt       # Python package requirements
├── 📦 package.json           # Node.js project configuration
└── 📖 README.md              # Project documentation

📝 Note: The chatterbox/ directory containing TTS models is generated during setup and not part of the repository.

🔌 API Reference

Endpoint	Method	Description
`/api/chat`	GET/POST	💬 Interactive chat functionality
`/api/generate`	POST	🎨 AI content generation services
`/api/images`	GET/POST	🖼️ Image processing and management
`/api/tts`	POST	🗣️ Text-to-speech conversion endpoint
`/api/videos`	GET/POST	🎬 Video content operations
`/api/voices`	GET/POST	🎵 Voice cloning and management

🔐 Environment Configuration

Create a .env.local file in the project root:

# Optional: Hugging Face authentication token for model access
HF_TOKEN=your_huggingface_token_here

🛠️ Development Workflow

🏃‍♂️ Available Commands

Command	Description
`npm run dev`	🚀 Start development server
`npm run build`	🔨 Create production build
`npm run start`	▶️ Start production server
`npm run lint`	🔍 Run ESLint code quality checks
`npm run setup`	⚙️ Complete environment setup