Z-Fusion

An enhanced Gradio interface for fast image generation powered by a ComfyUI backend.

This project is designed for 1-click installing via Pinokio

---

Screenshots

📸 Click to view screenshots

⚡ Image	✏️ Edit	🔍 Upscale	🧪 Experimental

---

⚡ Image

✏️ Edit

🔍 Upscale

🧪 Experimental

Features

• Z-Image - Image generation with the Z-Image Turbo and Base models
• FLUX2 Klein - Edit and Image gen workflows
• Txt2Img & Img2Img - with comfyui node parameters exposed to the UI
• SeedVR2 4K Upscaler - High-quality image and video upscaling
• LLM Prompt Assistants - Via Joycaption or in-built LLM support
• LoRA Support - Apply style/character Z-Image and Klein LoRAs
• GGUF friendly - Supports quantized models for low VRAM

License

MIT License - See LICENSE file

Third-Party

• ComfyUI - GPL-3.0
• ComfyKit - MIT
• Gradio - Apache-2.0

Credits

• Built with Gradio and ComfyKit
• Powered by ComfyUI
• GGUF support via ComfyUI-GGUF (Apache-2.0)
• Upscaling via ComfyUI-SeedVR2_VideoUpscaler (Apache-2.0)
• Video export via ComfyUI-VideoHelperSuite (GPL-3.0)
• Seed variance via SeedVarianceEnhancer (MIT)
• Wildcards based on comfyui_wildcards
• Camera prompts reference by CameraPromptsGenerator (MIT)
• Captioning via JoyCaption (GPL-3.0)