iamwavecut/telegram-chatgpt-bot

Simple Telegram bot integration to OpenAI ChatGPT API, written in Go

Simple Telegram bot integration to OpenAI ChatGPT API

---

Disclaimer

⚠️ This code represents a working instance of the bot named @net_nebot on Telegram. Other bots may be running the same code, and may be used for malicious purposes. Use at your own risk!

This is not an official OpenAI nor Telegram product. This is a community project.

The code is provided as is, and is not guaranteed to work.

---

Self-hosting

The easiest way to self-host the bot is to build a Docker image and run it on a server of your choice. The Dockerfile is provided in the repository.

You will need to provide the following arguments to the build command:

docker build -t telegram-chatgpt-bot . \
  --build-arg OPENAI_TOKEN=<required, your_api_token> \
  --build-arg BOT_TOKEN=<required, your_bot_token> \
  --build-arg CHAT_MODEL=<optional, defaults to 'gpt-3.5-turbo-1106'> 

You can also run the bot locally

docker run -d --restart always --name  telegram-chatgpt-bot telegram-chatgpt-bot

---

Translations

The bot is currently available in the following languages: English, Russian, Belarusian, Ukrainian. Feel free to contribute translations for other languages! The ChatGPT API itself understands a lot more languages, so go give it a try!

telegram-chatgpt-bot's Issues

Add queue and throttling to prevent rate limiting during surge

Control context length

Bot fails on context length error, if it's longer than 4096 tokens.

Should implement context trimmer.

Unescaped entities in requests

TBD: Needs to find out steps to reproduce.

400: Bad Request: can't parse entities: Can't find end of the entity starting at byte offset N