IOPaint

A free and open-source inpainting & outpainting tool powered by SOTA AI model.

📌 You are viewing the API Service branch

This branch provides a production-ready REST API service for watermark removal and inpainting.

• 🚀 Quick Start: See API_SERVICE_README.md
• 📖 API Documentation: See RESTFUL_API_DOCUMENTATION.md
• 🏗️ Deployment Guide: See API_SERVICE_GUIDE.md
• 📑 All Docs: See API_DOCS_INDEX.md

Looking for the WebUI version? Switch to main branch

Branch Comparison: See BRANCH_README.md

Erase(LaMa)	Replace Object(PowerPaint)

Draw Text(AnyText)	Out-painting(PowerPaint)

Features

• Completely free and open-source, fully self-hosted, support CPU & GPU & Apple Silicon

• Windows 1-Click Installer

• OptiClean: macOS & iOS App for object erase

• Supports various AI models to perform erase, inpainting or outpainting task.

  • Erase models: These models can be used to remove unwanted object, defect, watermarks, people from image.
  • Diffusion models: These models can be used to replace objects or perform outpainting. Some popular used models include:
    • runwayml/stable-diffusion-inpainting
    • diffusers/stable-diffusion-xl-1.0-inpainting-0.1
    • andregn/Realistic_Vision_V3.0-inpainting
    • Lykon/dreamshaper-8-inpainting
    • Sanster/anything-4.0-inpainting
    • BrushNet
    • PowerPaintV2
    • Sanster/AnyText
    • Fantasy-Studio/Paint-by-Example

• Plugins:

  • Segment Anything: Accurate and fast Interactive Object Segmentation
  • RemoveBG: Remove image background or generate masks for foreground objects
  • Anime Segmentation: Similar to RemoveBG, the model is specifically trained for anime images.
  • RealESRGAN: Super Resolution
  • GFPGAN: Face Restoration
  • RestoreFormer: Face Restoration

• FileManager: Browse your pictures conveniently and save them directly to the output directory.

Quick Start

Start webui

IOPaint provides a convenient webui for using the latest AI models to edit your images. You can install and start IOPaint easily by running following command:

# In order to use GPU, install cuda version of pytorch first.
# pip3 install torch==2.1.2 torchvision==0.16.2 --index-url https://download.pytorch.org/whl/cu118
# AMD GPU users, please utilize the following command, only works on linux, as pytorch is not yet supported on Windows with ROCm.
# pip3 install torch==2.1.2 torchvision==0.16.2 --index-url https://download.pytorch.org/whl/rocm5.6

pip3 install iopaint
iopaint start --model=lama --device=cpu --port=8080

That's it, you can start using IOPaint by visiting http://localhost:8080 in your web browser.

All models will be downloaded automatically at startup. If you want to change the download directory, you can add --model-dir. More documentation can be found here

You can see other supported models at here and how to use local sd ckpt/safetensors file at here.

Plugins

You can specify which plugins to use when starting the service, and you can view the commands to enable plugins by using iopaint start --help.

More demonstrations of the Plugin can be seen here

iopaint start --enable-interactive-seg --interactive-seg-device=cuda

Batch processing

You can also use IOPaint in the command line to batch process images:

iopaint run --model=lama --device=cpu \
--image=/path/to/image_folder \
--mask=/path/to/mask_folder \
--output=output_dir

--image is the folder containing input images, --mask is the folder containing corresponding mask images. When --mask is a path to a mask file, all images will be processed using this mask.

You can see more information about the available models and plugins supported by IOPaint below.

Model Recommendations

Choosing the right model depends on your use case and hardware. Here's our recommended model strategy:

🚀 Quick Start - For Daily Use

LaMa (Recommended for beginners)

iopaint start --model lama --device cuda --port 8080

• ⚡ Fastest - Near real-time processing
• 💾 Low VRAM - Uses ~1GB GPU memory
• 🎯 Best for: Removing watermarks, people, objects from images
• ✅ Most stable and reliable

🎨 Creative Editing - With Prompt Control

Stable Diffusion Inpainting

iopaint start --model runwayml/stable-diffusion-inpainting --device cuda --port 8080

• 🎨 Smart content generation - Not just removal, but intelligent filling
• 📝 Text prompts - Control what gets generated
• 🖼️ Creative flexibility - Replace objects with AI-generated content
• ✅ Official model - Well-maintained and stable

💎 Professional - High Quality Results

SDXL Inpainting (For high-resolution work)

iopaint start --model diffusers/stable-diffusion-xl-1.0-inpainting-0.1 --device cuda --low-mem --port 8080

• 🖼️ High resolution - Supports up to 1024x1024
• 🎨 Better details - Superior quality output
• 💎 Professional use - Best for photography and commercial work
• ⚠️ Requires more VRAM - Use --low-mem flag for optimization

📊 Model Comparison

Model	Speed	Quality	VRAM	Use Case	Recommended
LaMa	⚡⚡⚡⚡⚡	⭐⭐⭐⭐	~1GB	Quick erase	⭐⭐⭐⭐⭐
SD Inpainting	⚡⚡⚡	⭐⭐⭐⭐⭐	~4GB	Creative edit	⭐⭐⭐⭐⭐
SDXL Inpainting	⚡⚡	⭐⭐⭐⭐⭐	~8GB	Professional	⭐⭐⭐⭐
PowerPaint V2	⚡⚡⚡	⭐⭐⭐⭐	~5GB	Multi-task	⭐⭐⭐⭐

🔧 GPU Optimization Tips

For NVIDIA GPUs with limited VRAM:

# Enable low memory mode
iopaint start --model <model_name> --device cuda --low-mem --port 8080

# Enable CPU offload for very large models
iopaint start --model <model_name> --device cuda --cpu-offload --port 8080

For CPU-only systems:

# LaMa works well on CPU
iopaint start --model lama --device cpu --port 8080

📦 Installation Note

Updated Dependencies (2025-11-28)

This project now uses the latest stable versions of all dependencies. Install with:

# Recommended: Use mirror for faster installation (China users)
pip3 install -r requirements.txt -i https://mirrors.aliyun.com/pypi/simple/ --trusted-host mirrors.aliyun.com

# Or use official PyPI
pip3 install -r requirements.txt

See UPGRADE_NOTES.md for detailed information about package updates.

Development

Install nodejs, then install the frontend dependencies.

git clone https://github.com/Sanster/IOPaint.git
cd IOPaint/web_app
npm install
npm run build
cp -r dist/ ../iopaint/web_app

Create a .env.local file in web_app and fill in the backend IP and port.

VITE_BACKEND=http://127.0.0.1:8080

Start front-end development environment

npm run dev

Install back-end requirements and start backend service

pip install -r requirements.txt
python3 main.py start --model lama --port 8080

Then you can visit http://localhost:5173/ for development. The frontend code will automatically update after being modified, but the backend needs to restart the service after modifying the python code.