PyChatGPT is a Python client for ChatGPT that lets scripts hold real conversations with the chatbot — with conversation tracking, proxy support and session handling included. It became one of the most-starred ChatGPT client libraries of the early ChatGPT era and is still maintained in 2026. The project has about 4,200 stars on GitHub.
The library wraps ChatGPT in a small, Pythonic API: create a client, send a message, get the reply, and continue the same conversation with full history managed for you. It supports proxy configuration for restricted networks, token/session management, and both synchronous usage and integration into larger automation scripts. Developers used it to build Telegram and Discord bots, customer-support prototypes, data-labeling pipelines and personal assistants before official APIs covered every use case, and it remains handy where a ChatGPT Plus/Team-style conversational session — rather than a raw completion API — is what a workflow needs. The codebase is compact and dependency-light, so it is easy to audit and adapt; the repository continues to receive updates tracking changes on the ChatGPT side.
Historically, PyChatGPT arrived within weeks of ChatGPT’s launch and filled the gap before any official API existed, which is why its design centers on authenticated sessions rather than API keys. It handles the unglamorous parts of that approach — cookie/session persistence, retry behaviour and conversation branching — and its minimal surface area made it a common dependency in early ChatGPT bot frameworks. Users should be aware it automates a consumer product, so it suits personal and experimental projects best.
PyChatGPT is free and open source under the MIT license; you use it with your own ChatGPT account.
PyChatGPT is for Python developers automating workflows around ChatGPT conversations — bots, prototypes and internal tools — who want a ready-made client instead of hand-rolling session logic.