OpenAI on Thursday introduced a GitHub integration for the Deep Analysis instrument in ChatGPT, enabling customers to hyperlink repositories to ChatGPT. Based mostly on the queries, the deep analysis agent will retrieve pertinent data from the repository to compile stories.
For context, Deep Analysis on ChatGPT is a characteristic that performs complete analysis throughout a number of sources to craft prolonged stories. The characteristic was introduced final 12 months and is out there in numerous capacities throughout all ChatGPT plans.
Now you can join GitHub repos to deep analysis in ChatGPT.
Ask a query and the deep analysis agent will learn and search the repo’s supply code and PRs, returning an in depth report with citations. Hit deep analysis → GitHub to get began. pic.twitter.com/cdrQChfAU3— OpenAI Builders (@OpenAIDevs) Might 8, 2025
OpenAI additionally shared an instance for instance the characteristic’s capabilities. The report concerned shopping via OpenAI’s Codex CLI repository on GitHub. In a submit on X, Greg Brockman, president and co-founder of OpenAI, referred to as it “Deep Analysis to your codebase”.
GitHub is a platform that lets builders host code repositories for software program applications. It was acquired by Microsoft in 2018 for $7.5 billion. Microsoft has additionally actively backed OpenAI, with a number of investments. Nonetheless, a number of stories have hinted that the connection could have turned bitter.
In addition to, OpenAI releasing the characteristic signifies one other try at solidifying its worth proposition for builders. Just lately, it was reported that OpenAI reached an settlement to purchase Windsurf for $3 billion.
Final month, the corporate launched a brand new open supply code technology instrument referred to as Codex CLI—a light-weight coding agent that brings multimodal reasoning to the command line. Customers can mix screenshots or sketches with native code context, and the instrument is now open supply.
OpenAI additionally talked about that newly launched o3 and o4-mini fashions obtain greater scores in benchmarks reminiscent of Codeforces competitors code than their predecessors, citing a 20% fewer errors over OpenAI o1, as examined by exterior specialists.
“We’ve skilled them to make use of instruments, which isn’t one thing that we had performed with our earlier reasoning fashions. They really use these instruments of their chain of thought as they’re attempting to unravel a tough downside,” mentioned Brockman, including that in a single occasion, o3 used 600 instrument calls in a row attempting to unravel a very exhausting process.
The submit OpenAI’s GitHub Integration Brings ‘Deep Analysis for Your Code Base’ appeared first on Analytics India Journal.
