fooocus

Fooocus is a pre-configured AI image generation platform based on Stable Diffusion XL. It delivers a Midjourney-like experience — just type a prompt and get high-quality images. No complex workflows, no parameter tuning, no machine learning expertise required. This VM image comes with Fooocus, NVIDIA GPU drivers, Docker, and all dependencies pre-installed and ready to run.

Setting up AI image generation on cloud GPUs typically requires hours of manual configuration — installing NVIDIA drivers, CUDA, Python dependencies, and troubleshooting compatibility issues. Fooocus on Azure eliminates this entirely. Deploy the VM, open your browser, and start generating. No DevOps expertise needed.

Unlike subscription-based services like Midjourney, you maintain full control of your environment and data. Pay only for the GPU hours you use, with no per-image fees or monthly generation limits.

Fooocus is ideal for designers and creatives who need AI-generated images without learning complex tools, marketing teams producing visual content at scale, developers building AI-powered products who need a ready-to-deploy backend, and agencies offering AI art services who want a private, self-hosted solution.

Key Features

• Text-to-image with built-in GPT-2 prompt enhancement for beautiful results from simple prompts
• Image upscaling (1.5x and 2x) with AI-powered detail enhancement
• Inpainting and outpainting using Fooocus's custom algorithm
• Image prompts for style and composition guidance
• Face swap via InsightFace integration
• Multiple style presets including Default, Anime, and Realistic modes
• Web UI accessible on port 7865 — no SSH required

Getting Started

1. Deploy this image on a supported GPU VM (NC-series T4 recommended)
2. Open your browser to http://<vm-public-ip>:7865
3. Start generating images

Recommended VM Sizes

• Standard_NC4as_T4_v3 — 4 vCPUs, 28GB RAM, NVIDIA T4 16GB (best value)
• Standard_NC8as_T4_v3 — 8 vCPUs, 56GB RAM, NVIDIA T4 16GB