KetisBot

KetisBot is a powerful, voice-enabled AI chatbot for Discord. It leverages KoboldCpp for language generation, supports Stable Diffusion for image creation, performs web searches, and now offers voice interaction and text-to-speech playback in real-time.

Features

• 🤖 AI-generated responses using KoboldCpp
• 🎤 Voice interaction with speech recognition and TTS
• 🌐 Web search capability
• 🎨 Image generation via Stable Diffusion
• ⚙️ Customizable settings for admins
• 🛠 Slash commands for easy interaction

Installation

Prerequisites

• KoboldCpp
• Python 3.10+
• ffmpeg installed and in system PATH (required for voice and TTS)
• Discord bot token

Setup Instructions

1. Clone the repository:

   git clone https://github.com/Ketis21/KetisBot.git
   cd KetisBot

2. Install dependencies:

   pip install -r requirements.txt

3. Create a .env file in the project directory and add the following:

   BOT_TOKEN=your_discord_bot_token
   KAI_ENDPOINT=your_koboldcpp_api_url
   ADMIN_NAME=your_discord_username

4. Run the bot:

   python main.py

Commands

Interact with the bot by mentioning it (e.g., @KetisBot your message) or in the voice channel by saying Hey bot (e.g., Hey bot your message).

User Commands

• /search [query] – Search the internet
• /draw [orientation] [prompt] – Generate an image
• /describe [image] – Describe an uploaded image
• /reset – Clear chat history
• /joinvoice – Join your voice channel and listen
• /leavevoice – Leave the current voice channel

Admin Commands

• /maxlen [value] – Set response length (max 512)
• /idletime [value] – Set bot idle timeout
• /memory [text] – Override bot memory
• /settts [voice] – Change TTS voice

License

This project is licensed under the AGPL-3.0 license.

---