Forty-two Telegram bot that transcribes videos and audio messages to text using the OpenAI Whisper API

Send a video, audio message, or audio file to the bot, and it will transcribe it to text.

Setup

1. Create a .env file with the following content:

OPENAI_API_KEY=<your-openai-api-key>
TELEGRAM_API_TOKEN=<your-telegram-api-token>

2. Run the bot

With docker-compose:

docker-compose up -d

Run without Docker (ffmpeg is required):

poetry install
poetry run python main.py

Settings

Variable	Description	Default Value
OPENAI_API_KEY	OpenAI API key	-
TELEGRAM_API_TOKEN	Telegram API token	-
ALLOWED_USERS	List of allowed users (username or user id). Should be a Python list. Example: [1234567890, durov]	-

Obtaining API keys

• Telegram API key
• OpenAI API key

License

This project is licensed under the MIT License - see the LICENSE file for details.

Created with love in Barcelona