ChatGPT’s deep research tool gets a GitHub connector to answer questions about code

OpenAI improves its AI -driven “Deep Research” function with the possibility of analyzing code bases on Github.
On Thursday, OpenAi announced what it calls the first “connector” for Chatgpt Deep Research, the tool of the company looking for on the internet and other sources to put together thorough research reports on a subject. Now Chatgpt Deep Research can link to Github (in Beta), allowing developers to ask questions about codebase and engineering documents.
The connector will be available for Chatgpt Plus, Pro and Team users in the coming days, with Enterprise and EDU support, according to a OpenAI spokesperson.

The Github connector for Chatgpt Deep Research arrives when AI companies want to make their AI-driven chatbots more useful by building ways to link them to external platforms and services. Anthropic, for example, debuted integrations, which gives apps a pipeline in his AI Chatbot Claude.
Openi Years ago offered a plug-in options for chatgpt, but reduced it in favor of adapted chatbots called GPTS.
“I often hear that users find the deep research agent from Chatgpt so valuable that they want it to connect to their internal sources, beside the internet,” wrote OpenAI Head of Business Products Nate Gonzalez in a Blog post On LinkedIn. ‘[That’s why] Today we are introducing our first connector. “
In addition to answering questions about code bases, the new Chatgpt Deep Research Github Connector Chatgpt users break down product specifications in technical tasks and dependencies, summarize the code structure and patterns and understand how they can implement new APIs using Real Code examples.
WAN event
Berkeley, Ca
|
June 5
Book now
There is a risk that chatgpt deep research naturally hallucinates – of course there is no AI model that does not make confidence in it. But OpenAi throws the new possibility as a potential time savingser, no replacement for experts.
A openi spokesperson said Chatgpt will respect the settings of an organization, so that users only see Github content that they can already view and code bases that are explicitly shared with chatgpt.
OpenAi has invested in its tooling for auxiliary coding, recently reveals an open source coding tool for terminals called Codex CLI and upgrading the Chatgpt-Desktop app to read code in a handful of developer-oriented coding apps. The company sees programming as a top use for its models. An example of thisOpenAi reportedly reached an agreement to buy AI-driven coding assistant Windsurf for $ 3 billion.
In other OpenAI news on Thursday, the company launched the options for coordinating developers who want to adjust its newer models for certain applications. DEVS can now refine the O4-mini “reasoning model” of OpenAi through a technique that opened reinforcement of reinforcement, which uses task-specific assessment to improve the performance of the model. The refinement was also rolled out for the company’s GPT-4.1 Nano model.
Only verified organizations can refine O4-Mini, according to OpenAi. GPT-4.1 Nano-Verfijning is now available for all paying developers.
OpenAi started to mate certain models and developer functions behind verification, whereby organizations must submit an ID and other identity documents in April. The company claims that it is necessary to prevent abuse.