Gemini Discord Bot

A Discord bot that uses Google's Generative AI, capable of engaging in conversations and processing images.

Features

• Responds to mentions, direct messages, and when its name is used
• Maintains conversation context for each channel
• Processes and responds to images
• Slash commands for managing context:
  • Set context size
  • View current context
  • Clear context

Setup

1. Clone the repository

2. Install dependencies:

   npm install
   

3. Create a .env file with your Discord token and Google API key:

   DISCORD_TOKEN=your_discord_token_here
   GOOGLE_API_KEY=your_google_api_key_here
   

   • To create a Discord bot and get your token, visit: Discord Developer Portal

   • To obtain your Google API key for Gemini, go to: Google AI Studio

4. Configure the bot in config.ts:

   • Set BOT_NAME to your desired bot name
   • Choose the MODEL_NAME for the Google Generative AI model
   • Adjust SAFETY_SETTINGS to your preferred thresholds (For more information, see Gemini API Safety Settings)
   • Modify ALLOWED_IMAGE_TYPES if needed

5. Run the bot:

   npm start
   

Usage

• Mention the bot or use its name to start a conversation
• Send images for the bot to analyze and respond to
• Use slash commands to manage conversation context:
  • /view-context: Display the current context
  • /set-context: Set the number of messages to keep in context
  • /clear-context: Clear the current context

Dependencies

• discord.js
• @google/generative-ai
• dotenv
• undici