ChatGPT Telegram Bot

English | 中文

A Telegram bot with a silky smooth AI experience.

⚡Feature

[✓] Support for both Azure OpenAI and native OpenAI.

[✓] Real-time (streaming) response to AI, with faster and smoother experience.

[✓] 15 preset bot identities that can be quickly switched.

[✓] Support for custom bot identities to meet personalized needs.

[✓] Support to clear the contents of the chat with a single click, and restart the conversation at any time.

[✓]Native Telegram bot button support, making it easy and intuitive to implement required functions.

[✓] User level division, with different levels enjoying different single session token numbers, context numbers, and session frequencies.

[✓] Support English and Chinese on UI

[✓] Containerization.

[✓] More...

<p align="center"> <img src="https://media.giphy.com/media/gqKOf9LOL6xYK1Bmbv/giphy.gif" /> </p>

👨‍💻TODO

[x] Allow users to use their own OpenAI Key in the bot to gain more freedom.

[x] Improve ErrorHandler.

🤖Quick Experience

Telegram Bot: RoboAceBot

🛠️Deployment

Install Dependencies

pip install -r requirements.txt

Configure Database

Install Database

You can quickly create a local MySQL database using:

docker-compose up -d -f db/docker-compose.yaml

Initialize Database

mysql -uusername -p -e "source db/database.sql"

Add Configuration

All the required configurations are in config.yaml, please refer to config.yaml.example for file format and content.

Parameter	Optional	Description
BOT.TOKEN	No	Create a bot from @botFather and get the Token.
DEVELOPER_CHAT_ID	No	Telegram account ID that receives messages when the bot encounters an error. You can use @get_id_bot to get your ID.
MYSQL	No	Parameters related to MySQL connection.
TIME_SPAN	No	The time window size used to calculate the ratelimit, in minutes.
RATE_LIMIT	No	key is the user level, and value is the maximum number of chats that can be made within the TIME_SPAN time period.
CONTEXT_COUNT	No	key is the user level, and value is the number of contexts included in each chat.
MAX_TOKEN	No	key is the user level, and value is the maximum number of tokens returned by the AI per chat.
AI.TYPE	Yes	The type of AI used, with two options: openai and azure. The default is openai.
AI.BASE	Yes	When checking resources from the Azure portal, this value can be found in the "Keys and Endpoints" section. Alternatively, this value can be found in "Azure OpenAI Studio" > "Playground" > "Code View". Only needs to be set when AI.TYPE is azure.
AI.MODEL	Yes	The deployment name of Azure OpenAI, only needs to be set when AI.TYPE is azure.
AI.VERSION	Yes	The version number of Azure OpenAI, only needs to be set when AI.TYPEis azure.
AI.MODEL	Yes	The Model of OpenAI, only needs to be set when AI.TYPEis openai.

If you are using Azure's OpenAI, you can obtain all the required content at this link:

Get started using ChatGPT and GPT-4 with Azure OpenAI Services

🚀Start

python main.py | tee >> debug.log

Docker build & Run

docker run --rm --name chatgpt-telegram-bot -v ./config.yaml:/app/config.yaml ghcr.io/v-know/chatgpt-telegram-bot:latest 

Docker Compose

docker-compose up -d

❤️In Conclusion

I hope this project can provide you with a smooth AI experience and help more people create and use their own Telegram bots.