FastAPI AI Image Generation Service

A lightweight REST API that turns text prompts into images using diffusers/torch. Everything is wrapped in a Docker container for easy deployment anywhere that can run containers.

✨ Features

• /generate – returns a PNG file
  GET /generate?prompt=Two+cats+playing+chess
• /generate-memory – streams the PNG in‑memory (no temp file)
  GET /generate-memory?prompt=Cityscape+at+dusk
• Simple CRUD‑style sample routes (/, /items/{id}, /items/)

🐳 Quick start

# 1. Build the image
docker build -t ai-image-generator .

# 2. Run the container
docker run --rm -p 8000:8000 ai-image-generator

Visit http://localhost:8000/docs for the interactive Swagger UI.

🚀 Example calls

# Save an image to disk
curl -o output.png \
  "http://localhost:8000/generate?prompt=A+sunset+over+mountains"

# Stream an image
curl "http://localhost:8000/generate-memory?prompt=Futuristic+bike"

🛠 Local development

# Install dev tools
python -m venv .venv && source .venv/bin/activate
pip install -r dev-requirements.txt
pip install fastapi uvicorn[standard] pillow diffusers torch transformers

# Auto‑format, lint, and type‑check
black .               # formatting
ruff check .          # lint
mypy .                # type hints

Run locally:

uvicorn main:app --reload --port 8000

🏗️ Producing images

The heavy lifting happens inside ml.obtain_image(), which should accept:

image = obtain_image(
    prompt,
    num_inference_steps=50,
    seed=None,
    guidance_scale=7.5,
)

Make sure your implementation returns a PIL.Image instance.

📂 Project structure

.
├── main.py               # FastAPI routes
├── ml/                   # your ML / diffusion code
│   └── __init__.py
├── dev-requirements.txt  # lint, formatter, typing
└── Dockerfile

⚖️ License

MIT – do what you want, but no warranty is provided.
Remember to respect model licenses (e.g., Stable Diffusion) and content policies when deploying publicly.