GLM-Image

GLM-Image for Pinokio

A high-quality, professional-grade image generation UI powered by GLM-Image.

Run it locally with one click.

---

🚀 About

This application brings the power of GLM-Image to your desktop. GLM-Image is an open-source image generation model capable of producing stunning, high-resolution visuals from text prompts or existing images.

Built for Pinokio, this app offers a fully-featured Gradio interface that handles model downloading, environment setup, and inference automatically.

✨ Features

• 🎨 Text-to-Image Generation: Create detailed, high-resolution (up to 2048x2048) images from simple text descriptions.
• 🖼️ Image-to-Image Editing: Transform existing images using text prompts. Change styles, backgrounds, or details easily.
• 🎛️ Advanced Controls: Full control over generation parameters:
  • Resolution: Adjustable width and height (up to 2K).
  • Quality: Tune Inference Steps and Guidance Scale.
  • Seed Control: Randomize or lock seeds for reproducible results.
• 💾 Auto-Saving: All generated images are automatically saved to the output/ directory with timestamps.
• 🧹 Disk Space Optimizer: Includes a built-in "Save Disk Space" tool to deduplicate redundant library files, saving gigabytes of storage.

🛠️ System Requirements

• OS: Windows, Linux, or macOS
• GPU: NVIDIA GPU with 16GB+ VRAM recommended (The model is large ~33GB).
• Storage: At least 60GB of free disk space (for model weights and environment).

📦 How to Run

Option 1: One-Click Install (Pinokio)

1. Download and install Pinokio.
2. Navigate to the Discover page or paste this repository URL into the Pinokio browser:

   https://github.com/shinshekai/GLM-Image
   

3. Click Install.
4. Once installed, click Start.

Option 2: Manual Update / Reset

• Update: If a new version is released, click "Update" in the dashboard.
• Reset: If things break, use the "Reset" button to reinstall the environment (this won't delete your output/ images).
• Save Space: Click "Save Disk Space" in the dashboard menu to optimize storage usage.

⚖️ License

Software License: This application logic (UI and launcher scripts) is licensed under the terms found in the LICENSE file.

Model License: The GLM-Image model weights are released by Zhipu AI (zai-org) under the MIT License.

• Note: The tokenizer and specific weights derived from X-Omni-En are subject to Apache 2.0.
• Please verify your usage complies with the model's official license at Hugging Face.

🤝 Attribution

This project is built using:

• GLM-Image
• Gradio
• Diffusers
• PyTorch

---

_{Created with ❤️ by Shinshekai for the Pinokio Community}