Sumvoice

A powerful Telegram bot that transcribes and summarizes voice messages and audio files using Groq's AI models. Users can then ask questions about the transcribed content for interactive conversations.

🎯 Features

• Voice & Audio Transcription: Supports voice messages and audio files
• AI-Powered Summaries: Provides concise summaries in multiple languages (English, Russian, Ukrainian)
• Interactive Q&A: Ask questions about transcribed content
• Whitelist System: Optional user access control with admin management
• Multi-Language Support: Responds in the same language as the input
• Auto .env Creation: Automatically creates configuration files

🚀 Quick start

Prerequisites

• Python 3.7+
• Telegram Bot Token (from @BotFather)
• Groq API Key (from Groq Console)

Installation

1. Download ZIP or Git clone the repo

• ZIP download: https://github.com/LeeseTheFox/Sumvoice → Code → Download ZIP
• Git clone: git clone https://github.com/LeeseTheFox/Sumvoice.git

2. Install dependencies

   pip install -r requirements.txt

3. Run the bot

   python sumvoice_bot.py

4. Configure your tokens and admin

   • The bot will automatically create a .env file on first run
   • Edit the .env file and add your actual configuration:

   TELEGRAM_TOKEN=your_actual_telegram_token
   GROQ_API_KEY=your_actual_groq_api_key
   ADMIN_ID=your_telegram_user_id

5. Restart the bot

   python sumvoice_bot.py

⚙️ Configuration

Environment variables

The bot uses a .env file for configuration:

# Telegram Bot Configuration
TELEGRAM_TOKEN=your_telegram_token_here
GROQ_API_KEY=your_groq_api_key_here
ADMIN_ID=your_telegram_user_id_here

# Whitelist Configuration
WHITELIST_ENABLED=false
WHITELIST_IDS=123456789,987654321

Variable	Description	Default
TELEGRAM_TOKEN	Your Telegram bot token	Required
GROQ_API_KEY	Your Groq API key	Required
ADMIN_ID	Your Telegram User ID (for admin privileges)	Required
WHITELIST_ENABLED	Enable/disable user whitelist	false
WHITELIST_IDS	Comma-separated list of allowed user IDs	Empty

Finding your Telegram User ID

To set yourself as admin, you need your Telegram User ID:

1. Message @userinfobot on Telegram
2. Copy your User ID (numeric value)
3. Add it to your .env file:

   ADMIN_ID=123456789

📱 How to Use

For regular users

1. Send a voice message or audio file to the bot
2. Get an AI summary of the content
3. Ask questions about the transcribed content
4. Start fresh by sending a new voice/audio file

Example workflow

🎤 [Send voice message: "Today I learned about machine learning..."]
🤖 Bot: "AI summary: Discussion about machine learning concepts and applications."

💬 You: "What specific topics were mentioned?"
🤖 Bot: "The audio mentioned neural networks, supervised learning, and data preprocessing techniques."

💬 You: "Can you explain more about neural networks?"
🤖 Bot: "Based on the audio, neural networks were described as..."

🛠️ Available commands

User commands

Command	Description
/start	Start the bot and get welcome message
/help	Show help information

Admin commands

Command	Description
/whitelist [user_id]	Add a user to the whitelist
/toggle_whitelist	Enable/disable the whitelist system

🔒 Whitelist system

The bot includes an optional whitelist system:

• Disabled by default: Bot responds to everyone
• Easy toggle: Admins can enable/disable with /toggle_whitelist
• Persistent: Settings saved automatically to .env file
• Silent filtering: Non-whitelisted users are simply ignored

Managing the whitelist

1. Enable whitelist: /toggle_whitelist
2. Add users: /whitelist 123456789
3. Check status: /help (shows current state for admins)

🌍 Language support

The bot automatically detects and responds in:

• English
• Russian
• Ukrainian

For other languages, responses default to English.

🔧 Technical details

Dependencies

• python-telegram-bot - Telegram Bot API wrapper
• groq - Groq AI API client
• python-dotenv - Environment variable management
• requests - HTTP requests for audio processing

AI models used

• Transcription: whisper-large-v3-turbo
• Summarization & Q&A: meta-llama/llama-4-maverick-17b-128e-instruct

File processing

• Supports voice messages (.ogg format)
• Supports audio files (various formats)
• Temporary file handling with automatic cleanup
• Direct API integration with Groq's transcription service