This is a Telegram bot that lets you chat with the chatGPT over telegram. You can deploy and run the bot locally or via docker on a remote server. Prior implementations required running locally, in order to have the bot continuously online it is now able to be run anywhere.
You can view all features by sending /help to the bot
/ask, ask chatGPT anything receive a response/draw, draw pictures using stablediffusion/browse, give chatGPT access to Google/reload, force reload your session
Install miniconda and setup your environment
# install miniconda
wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh
chmod +x Miniconda3-latest-Linux-x86_64.sh
./Miniconda3-latest-Linux-x86_64.sh
# setup environment
conda env create -f environment.yml
playwright install
playwright install-deps
Build and run your Dockerfile
docker build . --tag chatgpt-telegram-bot
Copy .env.example and fill it in with the relevant API keys
cp .env.example .env
The .env file is layed out with the following keys
TELEGRAM_API_KEY=# (required) used to identify and control your bot
#TELEGRAM_USER_ID=# (optional) used to authenticate access to your bot to just a given account
OPENAI_EMAIL=# (required) email to access the openai account that is being used for chatgpt
OPENAI_PASSWORD=#(required) password to access the openai account that is being used for chatgpt
STABILITY_API_KEY=# (required) API key for stablediffusion to generate drawings
SERP_API_KEY=# (required) API key for google searches
You can obtain the required API Keys below
# if your environment is installed correct
python server.py
# if your environment is installed correct
mv telegram.service /etc/systemd/system/telegram.service
systemctl enable /etc/systemd/system/telegram.service
systemctl start telegram.service
docker run -d --name chatgpt-telegram-bot chatgpt-telegram-bot
You will also need to go the bot that you created via @Botfather, start a chat with it, and click start.
- Original Creator @Altryne on Twitter
- Based on Daniel Gross's whatsapp gpt package.