OpenAI is rolling out a GitHub connector for its ChatGPT “deep research” tool, enabling the agent to pull in and analyze private and public code repositories. Over the next few days, Plus, Pro, and Team subscribers will be able to ask the model questions about codebases and documentation they’ve shared, with Enterprise and Education tiers to follow. The integration can break down product specs into actionable technical tasks, summarize code structure and patterns, and illustrate API usage with real examples—while respecting each organization’s existing repository permissions.
This launch comes as part of a broader push to link ChatGPT to external platforms—reminiscent of Anthropic’s new Claude integrations—and follows OpenAI’s deprecation of its earlier plug-in system in favor of custom GPTs. Alongside the GitHub connector, OpenAI also announced new fine-tuning options: verified organizations can now apply reinforcement fine-tuning to the reasoning-focused o4-mini model, and all paying developers can fine-tune the GPT-4.1 nano. These gated capabilities, introduced in April, require identity verification to help guard against abuse.