ChatGPT adds GitHub connector to analyse codebases

Feature available to ChatGPT Plus, Pro, Team users, with support for Enterprise and Education tiers coming soon

Listen to article

OpenAI has introduced a GitHub connector for its ChatGPT deep research tool, enabling users to ask questions about codebases and engineering documents directly from GitHub repositories.

The feature, launching in beta, is initially available to ChatGPT Plus, Pro, and Team users, with support for Enterprise and Education tiers coming soon.

This marks the first official connector for ChatGPT deep research, a feature designed to compile comprehensive web-based reports on technical and non-technical topics.

The GitHub integration allows developers to break down product specs into tasks, summarize code structure and patterns, and understand API implementations using real code.

OpenAI Head of Business Products Nate Gonzalez attributed user demand for internal data integration, stating the connector is a direct response to developer feedback.

While the AI may still “hallucinate” answers, OpenAI positions the feature as a time-saving assistant rather than a replacement for expert review.

To safeguard sensitive data, the tool respects GitHub permission settings—only accessing content explicitly shared with ChatGPT.

The launch comes amid increased investment by OpenAI in developer tooling.

Recent efforts include the release of Codex CLI, an open-source terminal coding assistant, and upgrades to the ChatGPT desktop app for improved code readability. OpenAI is also reportedly acquiring Windsurf, an AI coding assistant, for $3 billion.

In parallel, OpenAI rolled out new fine-tuning capabilities for developers. Its o4-mini “reasoning” model now supports reinforcement fine-tuning, while the GPT-4.1 nano model is also customizable.

Only verified organizations can fine-tune o4-mini, part of OpenAI’s April shift toward identity-gated access to prevent misuse.

As AI chat tools expand into code comprehension, OpenAI’s GitHub connector signals a broader strategy to embed generative AI into real-world software workflows.