GPT-4 Telegram Chatbot

This script allows you to create a Telegram chatbot that responds to user messages using the GPT-4 language model. The chatbot works by fetching the last X messages in a conversation and generating a response based on the conversation history.

Demo: http://t.me/rvnikita_public

Author blog: http://t.me/rvnikita_blog

How to generate a session string

To avoid confirming your account with a code every time the script is launched, you can generate a session string and store it as an environment variable. To generate the session string, run the generate_session_string() function in the script.

Environment Variables

The following environment variables are required for the script to work:

API_ID: Your Telegram API ID API_HASH: Your Telegram API Hash PHONE_NUMBER: Your Telegram account phone number (e.g., +1234567890) OPENAI_API_KEY: Your OpenAI API Key (make sure you have access to the GPT-4 model or swith to the GPT-3.5 model) TELEGRAM_SESSION_STRING: The session string generated using the generate_session_string() function

How to use

Install the required dependencies:

pip install -r requirements.txt

Set the environment variables with the appropriate values:

export API_ID=your_api_id
export API_HASH=your_api_hash
export PHONE_NUMBER=your_phone_number
export OPENAI_API_KEY=your_openai_api_key
export TELEGRAM_SESSION_STRING=your_telegram_session_string
exporr LOGGING_CHAT_ID=your_logging_chat_id

Run the dispatcher.py script:

python dispatcher.py

The script will connect to your Telegram account and start listening for new messages. When a new message is received, it will generate a response using the GPT-4 language model and send it to the user.