This ChatGPT Telegram Bot, maintained by developer yym68686 and also known as TeleChat, turns Telegram into a full AI chat client backed by large language models. Hosted on GitHub with about 1,300 stars under the GPL-3.0 license and actively updated through 2026, it is a distinct, community-run project — not the similarly named bot products listed elsewhere.
The bot connects Telegram chats to OpenAI-compatible LLM APIs, so you can hold conversations and run web searches with frontier models without leaving the messenger. According to the project, the bot has been used with models such as GPT-5, DALL·E for images, Groq-hosted models, Gemini and the official Claude API, and it can sit behind aggregators like one-api/new-api or the author’s companion uni-api project to reach providers including Anthropic, Vertex AI, Azure, AWS and OpenRouter. Deployment is flexible: ready-made paths cover Docker (with a published Docker Hub image), Zeabur, fly.io, Koyeb and Replit, so the bot can be running on free or cheap hosting within minutes using your own API keys.
The bot is free and open source under the GPL-3.0 license. You host it yourself — often on free tiers — and pay only for the model API usage through your own keys with whichever provider you configure.
Telegram power users and small communities who want a capable multi-model AI assistant inside their chats, and developers looking for a proven, deployable Telegram bot codebase to self-host or extend.