Neura Market
chatgpt-telegram-bot logo

chatgpt-telegram-bot

Free

Telegram bot that supports voice messages.

FreeFree tier
Inputs: text, image, audioOutputs: text, image
Type
Open Source
Visit Website GitHub (40 stars)

About chatgpt-telegram-bot

ChatGPT Telegram Bot is an open-source Python-based project that re-creates the ChatGPT experience as a Telegram bot. It integrates with OpenAI's API and Anthropic Claude models via OpenRouter, allowing users to interact with frontier AI models directly within Telegram. The bot supports low-latency replies, message streaming, and has no daily request limits when users provide their own API keys. It is designed for both individual and group chat use, with 15 special chat modes including Assistant, Code Assistant, Psychologist, and more, which can be customized in the configuration file.

The bot offers a range of multimodal capabilities: it can process voice messages by transcribing them with Whisper, accept images for vision-based analysis using models like GPT-5.5 or Claude, and generate images via OpenAI's gpt-image-1 model in an Artist mode. It also supports group chats and can be deployed locally using Docker or a simple two-command setup. The project is maintained on GitHub under the father-bot/chatgpt_telegram_bot repository and includes a live demo bot (@jadvebot) and a web interface (jadve.com).

Key Features

Low latency replies (typically 3-5 seconds)
No daily request limits when using own API keys
Message streaming for word-by-word responses
Support for GPT-5.5 and Anthropic Claude models (Opus, Sonnet, Haiku) via OpenRouter
Vision capability: send images for analysis by vision-capable models
Image generation via OpenAI gpt-image-1 in Artist mode
Voice message transcription using Whisper
15 customizable chat modes (Assistant, Code Assistant, Psychologist, etc.)

Pros & Cons

Pros
  • Open-source and self-hostable, giving users full control over their data and API usage
  • No built-in daily request limits when using personal API keys
  • Supports multiple frontier models (GPT-5.5, Claude) via OpenRouter
  • Multimodal: handles text, voice, images, and image generation
  • Low latency and message streaming for a responsive chat experience
  • Customizable chat modes and configuration
Cons
  • Requires users to obtain and configure their own API keys from OpenAI or OpenRouter
  • Free tier of the live bot may have usage limits; self-hosting is recommended for unlimited use
  • Setup requires some technical knowledge (Docker, Python, or command-line skills)
  • Voice transcription and image generation depend on external API availability and costs
  • Group chat setup may require additional configuration steps

Best For

Getting quick AI-powered answers and assistance directly in TelegramTranscribing voice notes into text using WhisperAnalyzing images with vision-capable AI modelsGenerating images from text prompts within TelegramUsing specialized chat modes for coding help, psychological advice, or role-playingDeploying a private AI bot for group chats with no usage limits

FAQ

Do I need my own API key to use this bot?
Yes, the bot uses your own API keys from OpenAI or OpenRouter. The live demo bot (@jadvebot) may have its own limits, but for full control and no daily limits, you should deploy your own instance with your keys.
Can the bot handle voice messages?
Yes, the bot supports voice messages. It uses Whisper to transcribe voice notes into text, which can then be processed by the AI models.
Does the bot support image generation?
Yes, the bot includes an Artist mode that uses OpenAI's gpt-image-1 model to generate images from text prompts.
Is the bot free to use?
The bot software is open-source and free to deploy. However, using the live demo bot may involve usage limits, and you will incur API costs from OpenAI or OpenRouter when using your own keys.
Can I use the bot in group chats?
Yes, the bot supports group chats. You can run /help_group_chat in the bot for setup instructions.
What models does the bot support?
Based on available information, the bot supports GPT-5.5 and Anthropic Claude models (Opus, Sonnet, Haiku) via OpenRouter, as well as OpenAI's gpt-image-1 for image generation.