GPT Code UI

GPT Code UI is an open-source implementation of OpenAI’s ChatGPT Code Interpreter: a web interface where you upload files, describe what you want, and an LLM writes and runs Python in a local Jupyter kernel to analyse data, build charts and produce downloadable results. Released when the official Code Interpreter was still gated, it gave anyone with an API key the same workflow. The project has around 3,500 stars on GitHub and is MIT licensed.

Overview

The app pairs a React front end with a small Python server that manages a Jupyter kernel per session. When you ask a question, the model generates code, the server executes it in the kernel, and outputs — text, plots, tables, files — stream back into the chat. Because execution happens in a persistent kernel, variables and dataframes survive between messages, so analysis is genuinely iterative: clean the data, plot it, refine the plot, export the result. It works with GPT-4 and GPT-3.5-class models via the OpenAI API, and everything runs locally, which keeps uploaded files on your own machine.

Key Features

• Chat interface over a persistent Jupyter kernel
• Upload CSVs and other files for interactive data analysis
• Generated code is executed and its outputs rendered inline
• Charts, tables and generated files returned in the conversation
• Runs entirely locally apart from the model API calls
• Simple Docker-friendly deployment
• MIT licensed

💳 Pricing

GPT Code UI is completely free and open source under the MIT license. You only pay for your own OpenAI API usage.

Who Is It For

Analysts and developers who want a self-hosted Code Interpreter experience — chatting with their data and getting real executed results — without depending on a ChatGPT subscription feature.