Getting Started with Fooocus-API: Docker Setup and First API Calls
A step-by-step tutorial for setting up Fooocus-API with Docker and making your first AI image generation API calls.
Getting Started with Fooocus-API: Docker Setup and First API Calls
This tutorial walks through setting up Fooocus-API with Docker and making your first image generation requests.
Prerequisites
- A machine with an NVIDIA GPU - Fooocus-API requires CUDA
- Docker installed
- NVIDIA Container Toolkit for GPU passthrough
- 16GB+ RAM and 20GB free disk for model downloads
Step 1: Pull and Run the Docker Container
docker run -d --gpus=all \
-e NVIDIA_DRIVER_CAPABILITIES=compute,utility \
-e NVIDIA_VISIBLE_DEVICES=all \
-p 8888:8888 \
--name fooocus-api \
konieshadow/fooocus-api:latest
For persistent model caching:
docker run -d --gpus=all \
-e NVIDIA_DRIVER_CAPABILITIES=compute,utility \
-e NVIDIA_VISIBLE_DEVICES=all \
-v ~/repositories:/app/repositories \
-v ~/.cache/pip:/root/.cache/pip \
-p 8888:8888 \
--name fooocus-api \
konieshadow/fooocus-api:latest
Step 2: Wait for Model Downloads
On first startup, models download automatically (5-15 minutes). Watch logs:
docker logs -f fooocus-api
Step 3: Generate Your First Image
curl -X POST "http://localhost:8888/v1/generate" \
-H "Content-Type: application/json" \
-d '{
"prompt": "a serene mountain lake at sunset, cinematic lighting",
"width": 1024,
"height": 1024,
"guidance_scale": 7,
"steps": 30
}'
Poll for completion with your task ID:
curl "http://localhost:8888/v1/generation/result?task_id=YOUR_TASK_ID"
Using Docker Compose
services:
fooocus-api:
image: konieshadow/fooocus-api:latest
restart: unless-stopped
ports:
- "8888:8888"
volumes:
- ./repositories:/app/repositories
- ./pip-cache:/root/.cache/pip
deploy:
resources:
reservations:
devices:
- driver: nvidia
count: 1
capabilities: [gpu]
Conclusion
Fooocus-API makes it straightforward to integrate professional-quality AI image generation into any application. With Docker, you can be up and running in minutes, generating production-ready images through a clean REST API.
#fooocus
#docker
#tutorial
#image-generation
#stable-diffusion