Gemini ImageGen

Simple windowed desktop app that uses a Gemini-like API to generate images from a user prompt and display them with a "Save as PNG" option. The UI uses a dark theme and scales the displayed image to the window.

Prerequisites

• Python 3.11+
• Install dependencies from requirements.txt (see below)
• Optionally set GEMINI_API_KEY and GEMINI_ENDPOINT environment variables for a real API.

Install

Open PowerShell and run:

python -m pip install -r requirements.txt

Running

To run the app:

# From the project folder
python main.py

Environment variables (optional)

• GEMINI_API_KEY : Your API key (if omitted, the app shows a placeholder image locally).
• GEMINI_ENDPOINT : The image generation endpoint (defaults to a placeholder URL in api.py).

PowerShell example to set the key for the session:

$env:GEMINI_API_KEY = "YOUR_KEY_HERE"
python main.py

Notes for integrating a real Gemini endpoint

The api.generate_image function shows how to send a JSON payload with a prompt field and handles two common response shapes:

• image_base64 (base64 string containing image bytes) or
• image_url (URL pointing to the generated image)

Adjust api.py to match the exact Gemini request and response format your account expects.

Files

• main.py — app entry
• gui.py — Qt UI and threading
• api.py — Gemini adapter + placeholder image generator
• requirements.txt — dependencies