ChatGPT’s Deep Research Feature Can Now Connect With GitHub Repositories

ChatGPT is getting a new feature that will enable users to connect their GitHub repository with the chatbot. Announced last week, the new feature is part of the Deep Research agent. OpenAI said the new feature will allow the artificial intelligence (AI) chatbot to prepare comprehensive reports based on the user's and others' repositories. Currently, the San Francisco-based AI firm is rolling out the feature to only the paid subscribers of the platform. Notably, it can take up to five minutes before a newly created or edited repository begins showing up in the Deep Research GitHub connector.

GitHub Connector Is Now Part of Deep Research in ChatGPT

In a post on X (formerly known as Twitter), the official handle of OpenAI Developers announced the rollout of the new feature. ChatGPT's Deep Research with GitHub is available globally to Team users. All Plus and Pro subscribers will also get the feature, except those in the EEA, Switzerland, and the UK.

The account also shared a video demo of the feature. Based on that, the eligible users will first have to ask Deep Research a query. Then they will see a new GitHub connection button, tapping which will give the chatbot access to the user's, as well as any other third-party public repository.

“Ask a question and the deep research agent will read and search the repo's source code and PRs, returning a detailed report with citations,” OpenAI said.

In its support pages, the AI firm highlighted that the feature is currently in beta, and GitHub's terms and conditions will apply. When Deep Research connects to GitHub, it lets ChatGPT pull live data from repositories, including code, Readme files, and other documents. The chatbot can also analyse the data to provide insights about the content.

Notably, OpenAI highlights that if ChatGPT Plus and Pro users have kept the “Improve the model for everyone” setting on, the company will use the GitHub data to train its models. However, if the setting is turned off, the data will not be stored. The AI firm added that it does not store data from its business offerings, including Team, Enterprise, Edu, and its application programming interface (API).