GPT-Telegram-Bot: Multifunctional AI Assistant 🤖💬

English | 简体中文 | 繁體中文 | 日本語 | Español | Français | Русский | Deutsch

GPT-Telegram-Bot is a powerful Telegram bot that integrates various AI models, providing intelligent conversations, image generation and analysis capabilities.

Main Features 🌟

1. Multi-model Support 🎭: Any model compatible with OpenAI API, Google Gemini, Anthropic Claude, Groq, and Azure OpenAI
2. Intelligent Conversations 💬: Natural language interactions with context memory support
3. Image Generation 🎨: Create images based on text descriptions
4. Image Analysis 🔍: Interpret and describe uploaded images
5. Multilingual Support 🌐: Localization support for multiple languages
6. Stream Response ⚡: Real-time generation and display of AI replies
7. User Whitelist 🔐: Can be set to allow access only to specific users

Supported AI Models 🧠

• OpenAI Series: Powerful language models 🚀
• Google Gemini: Google's next-generation AI model 🧑‍🔬
• Anthropic Claude: Another powerful language model choice 🎭
• Groq: High-speed inference AI model ⚡
• Azure OpenAI: Microsoft-hosted OpenAI service 👔

Project Structure 📁

GPT-Telegram-Bot/
├── api/                # API related configuration
│   ├── telegram.js     # Handles Telegram bot interactions
├── src/                # Source code
│   ├── api.js          # Handles interactions with OpenAI compatible APIs 
│   ├── bot.js          # Main Telegram bot logic
│   ├── config.js       # Configuration file
│   ├── azureOpenAI.js  # Handles interactions with Azure OpenAI
│   ├── claude.js       # Handles interactions with Claude
│   ├── generateImage.js# Handles interactions with DALL·E
│   ├── geminiApi.js    # Handles interactions with Gemini
│   ├── groqapi.js      # Handles interactions with Groq
│   ├── uploadhandler.js# Handles image upload and analysis logic
│   ├── localization.js # Handles multilingual support
│   ├── redis.js        # Upstash Redis database functionality
├── locales/            # Multilingual support files
│   ├── en.json         
│   ├── zh-cn.json      
│   ├── zh-hant.json    
│   └── ja.json         
│   └── es.json         
│   └── fr.json         
│   └── ru.json         
│   └── de.json         
├── docs/               # Multilingual README files
│   ├── README.zh-cn.md 
│   ├── README.zh-hant.md
│   ├── README.ja.md    
│   ├── README.es.md    
│   ├── README.fr.md    
│   ├── README.ru.md    
│   └── README.de.md    
├── public/             # Vercel webpage after deployment
│   └── index.html      # Webpage entry file
├── package.json        # Project dependencies
├── vercel.json         # Vercel configuration file
└── README.md           # Project description file

Quick Start 🚀

Prerequisites

• Vercel account
• Telegram account and Bot Token
• Upstash Please select the Redis database and enable the Eviction feature
• API key for at least one AI service

Deployment Steps

1. Clone the repository:

   git clone https://github.com/snakeying/GPT-Telegram-Bot.git
   cd GPT-Telegram-Bot

2. Install dependencies:

   npm install

3. Configure environment variables: Create a .env file and fill in the necessary configuration information (refer to the environment variable configuration below).

4. Deploy to Vercel:

   • Fork this repo
   • Modify according to the instructions at the bottom of the readme
   • Click the "Deploy with Vercel" button
   • Connect your GitHub repository
   • Configure environment variables
   • Complete deployment

5. Set up Telegram Webhook: After deployment, use the following URL to set up the Webhook:

   https://api.telegram.org/bot<YOUR_BOT_TOKEN>/setWebhook?url=<YOUR_VERCEL_DOMAIN>/api/telegram
   

Environment Variable Configuration 🔧

Before deploying and running GPT-Telegram-Bot, you need to set the following environment variables. Create a .env file in the project root directory and configure the following variables:

Variable Name	Description	Default Value
OPENAI_API_KEY	OpenAI API key	-
OPENAI_BASE_URL	OpenAI API base URL	https://api.openai.com/v1
OPENAI_MODELS	OpenAI models to use (comma-separated)	-
DEFAULT_MODEL	Default model to use	First model in OPENAI_MODELS
AZURE_OPENAI_API_KEY	Azure OpenAI API key	-
AZURE_OPENAI_ENDPOINT	Azure OpenAI endpoint	-
AZURE_OPENAI_MODELS	Azure OpenAI models to use (comma-separated)	-
TELEGRAM_BOT_TOKEN	Telegram Bot Token	-
WHITELISTED_USERS	Allowed user IDs (comma-separated)	-
DALL_E_MODEL	DALL-E model to use	dall-e-3
UPSTASH_REDIS_REST_URL	Upstash Redis REST URL	-
UPSTASH_REST_TOKEN	Upstash Redis REST Token	-
SYSTEM_INIT_MESSAGE	System initialization message	You are a helpful assistant.
SYSTEM_INIT_MESSAGE_ROLE	System message role	system
GEMINI_API_KEY	Google Gemini API key	-
GOOGLE_MODELS	Google models to use (comma-separated)	-
GEMINI_ENDPOINT	Gemini API endpoint	https://generativelanguage.googleapis.com/v1beta/models
GROQ_API_KEY	Groq API key	-
GROQ_MODELS	Groq models to use (comma-separated)	-
MAX_HISTORY_LENGTH	Maximum history length	50
CLAUDE_API_KEY	Anthropic Claude API key	-
CLAUDE_MODELS	Claude models to use (comma-separated)	-
CLAUDE_ENDPOINT	Claude API endpoint	https://api.anthropic.com/v1/chat/completions

Make sure to add these environment variables to your project's environment configuration when deploying to Vercel or other platforms.

Usage Guide 📖

• /start - Initialize the bot
• /new - Start a new conversation
• /history - View conversation history summary
• /help - Get help information
• /switchmodel <model name> - Switch AI model
• /img <description> [size] - Generate image
• /language <language code> - Switch interface language
• Send an image for analysis
• Send a message directly for conversation

Supported languages (use /language command):

• English (en)
• Simplified Chinese (zh-cn)
• Traditional Chinese (zh-hant)
• Japanese (ja)
• Spanish (es)
• French (fr)
• Russian (ru)
• German (de)

Notes ⚠️

• Use API quotas reasonably, especially when using image features 💸
• Securely store environment variables and API keys 🔒
• Different AI models may have different features and limitations 🔄
• Regularly check and update dependencies to ensure security and performance 🔧

Contribution 🤝

Welcome to submit Pull Requests or open Issues to improve this project! Your contributions will make this AI assistant more powerful and interesting.

License 📜

This project is licensed under the MIT License.

---

About the "Deploy to Vercel" button: This button provides a one-click deployment to Vercel, which is very convenient. However, please note:

1. The link in the button points to the original repository (https://github.com/snakeying/GPT-Telegram-Bot).
2. If you've forked this project and want to deploy your own version, you need to update this button link in the README.
3. Update method: Replace snakeying/GPT-Telegram-Bot in the link with your GitHub username and repository name.

For example, if your GitHub username is "yourname", you should change the button link to:

[![Deploy with Vercel](https://vercel.com/button)](https://vercel.com/new/clone?repository-url=https%3A%2F%2Fgithub.com%2Fyourname%2FGPT-Telegram-Bot)

This ensures that the "Deploy to Vercel" button will deploy your forked version, not the original repository.