Red_STT_Bot

A Telegram bot for speech recognition from voice and video messages using Nexara (OpenAI Whisper).

Features

• Accepts voice and video messages in Telegram.
• Recognizes speech using the Nexara STT API (based on OpenAI Whisper).
• Sends the user a text transcription of the audio.
• Logs events via Logfire.
• Can work in groups: automatically recognizes speech from voice and video messages in a group (if given admin rights).

Quick Start

1. Clone the repository

git clone https://github.com/BuT9l3b/Red_STT_Bot.git
cd Red_STT_Bot

2. Install dependencies

pip install -r requirements.txt

3. Set environment variables

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

TG_TOKEN=your_telegram_token
NEXARA_TOKEN=your_nexara_token
LOGFIRE_TOKEN=your_logfire_token

4. Run the bot

python main.py

Project Structure

• main.py — main file to run the Telegram bot.
• stt.py — module for working with the Nexara STT API.
• message_tools.py — helper functions for sending messages.
• texts.py — text templates for messages.

Usage

1. Start the bot.
2. Send /start to it in Telegram.
3. Send a voice or video message.
4. Receive the text transcription of the audio.

Dependencies

• Python 3.8+
• aiogram
• openai
• httpx
• python-dotenv
• logfire
• telegramify_markdown

Install all dependencies with:

pip install -r requirements.txt

License

MIT License