🖼️ Pollinations Image Proxy API

A lightweight proxy service compatible with the OpenAI /v1/images/generations endpoint. It forwards OpenAI-style image generation requests to Pollinations.

🚀 Features

• ✅ OpenAI-style Compatibility
• 🔐 Requires Authorization: Bearer pollinations
• 🐳 Docker & non-Docker ready
• ⚡ Built with FastAPI

📦 Project Structure

.
├── main.py
├── requirements.txt
├── Dockerfile
└── README.md

🛠️ Quick Start

🔧 Run Without Docker

pip install -r requirements.txt
uvicorn main:app --host 0.0.0.0 --port 5000

🐳 Run with Docker

docker build -t pollinations-api .
docker run -d -p 5000:5000 pollinations-api

🔑 API Key

All requests must include the following header:

Authorization: Bearer pollinations

📡 API

POST /v1/images/generations

Request Body

{
  "prompt": "a futuristic city with flying cars",
  "width": 512,
  "height": 512,
  "model": "flux",
  "enhance": true
}

Parameters

Name	Type	Default	Description
prompt	string	—	Text description of the image to generate
width	integer	512	Image width in pixels
height	integer	512	Image height in pixels
model	string	flux	The AI model to employ for image generation (see below)
enhance	boolean	false	Apply high-resolution enhancement to the generated image

Model Options

The model field influences both quality and generation style:

• flux: General purpose, good quality (default)
• flux-pro: General, best quality but slower
• flux-realism: Realistically styled images without artistic embellishments
• flux-anime: Ideal for anime, manga, and comic styles
• flux-3d: Best for intricate 3D-like scenes
• flux-cablyai: General purpose with Cablyai inference engine
• turbo: Medium to low quality, but marginally faster

Choose a task-specific model if your prompt is closely aligned with one of these styles. Otherwise, flux is a solid general-purpose choice.

📄 License

This project is licensed under the MIT License. See the LICENSE file for details.