🍌 Playing with Nano Banana on Discord Together 🍌

繁體中文 | English

Overview

Nano Banana is a tutorial project demonstrating how to use Google Gemini API for AI image generation. Through a Discord bot interface, users can:

• 🎨 Text-to-Image - Create images from text descriptions
• 🔄 Image Transformation - Modify existing images based on prompts
• 💬 Discord Integration - Interact via slash commands or messages

Technical Highlights

• Clean modular architecture (config, API, Discord layers separated)
• Full async programming implementation
• Configuration management with Pydantic
• Complete test coverage (pytest + pytest-asyncio)
• Fast dependency management with uv

Quick Start

Prerequisites

• Python 3.13+
• Google Gemini API Key
• Discord Bot Token

Installation

# Install uv package manager
pip install uv

# Clone the repository
git clone https://github.com/GDG-NTUST/course-1141-9.git
cd course-1141-9

# Install dependencies
uv sync

Configuration

Create a .env file:

GOOGLE_API_KEY=your_google_api_key
DISCORD_TOKEN=your_discord_bot_token
DISCORD_GUILD_ID=your_server_id

# Optional settings
LOG_LEVEL=INFO
MODEL_NAME=gemini-2.5-flash-image
SYSTEM_PROMPT=You are a helpful AI assistant.

Running

# Start the Discord bot
cd src/
uv run nano_banana

Usage

Discord Bot

Slash Command:

/draw a cat wearing a hat

Message Interaction:

• Send text messages directly to generate images
• Attach images with text prompts for image transformation
• Reply to messages containing images

Command-Line Demo

cd src/

# Text-to-Image
uv run nano_banana_cli -p "beautiful sunset"

# Image Transformation
uv run nano_banana_cli -p "make it more vibrant" -i image.png

# Multiple Images
uv run nano_banana_cli -p "combine these styles" -i img1.png img2.png

Generated images are saved in src/nano_banana/api/demo/outputs/

Project Architecture

nano_banana/
├── core/
│   └── config.py          # Pydantic configuration management
├── api/
│   ├── client.py          # Google Gemini API client
│   └── demo/
│       └── demo.py        # Command-line demo
└── discord/
    ├── bot.py             # Discord bot
    └── utils.py           # Utility functions

Core Modules

• config.py - Environment variable loading, logging configuration, settings validation
• client.py - Async Gemini API client supporting text-to-image and image transformation
• bot.py - Discord bot handling slash commands and message events
• utils.py - Image download and response handling utilities

Testing

# Install development dependencies
uv sync

# Run all tests
uv run pytest

# Run tests with coverage report
uv run pytest --cov=nano_banana --cov-report=html

# Run specific tests
uv run pytest tests/api/test_client.py

Test Coverage:

• ✅ Configuration management (environment variables, validation, defaults)
• ✅ API client (text-to-image, image transformation, error handling)
• ✅ Discord bot (command handling, message listening, image attachments)
• ✅ Utility functions (image download, response formatting)

Development

Code Quality

# Format code
uv run ruff format .

# Run linter
uv run ruff check . --fix

Project Configuration

• Ruff - Code linting and formatting
• pytest - Testing framework
• pytest-asyncio - Async testing support
• pytest-cov - Test coverage

See pyproject.toml for detailed configuration

Supported Models

Model Name	Max Images
gemini-2.5-flash-image	3
gemini-3-pro-image-preview	14

Set MODEL_NAME in .env to switch models.

Troubleshooting

API Key Error:

• Verify .env file exists and is properly formatted
• Check if API key is valid and has quota

Discord Bot Not Responding:

• Confirm bot is online
• Check if Discord Token is correct
• Ensure Message Content Intent is enabled
• Verify DISCORD_GUILD_ID is set correctly

Enable Debug Logging:

LOG_LEVEL=DEBUG

License

This project is licensed under the MIT License

Contributing

Pull requests and issues are welcome!

---

Made with ❤️ by GDG NTUST