OpenAI is enhancing its AI-powered "Deep Research" feature by adding the ability to analyze GitHub repositories.
On Thursday, OpenAI announced the launch of its first so-called "connector." ChatGPT Deep Research is a tool from the company that can search the web and other sources to create detailed research reports on a given topic. Now, ChatGPT Deep Research can connect to GitHub (beta), allowing developers to ask questions about codebases and engineering documents.
According to an OpenAI spokesperson, this connector will be available to ChatGPT Plus, Pro, and Team users in the coming days, with enterprise and educational support also set to launch soon.
The GitHub connector for ChatGPT Deep Research comes as AI companies seek to make their AI-driven chatbots more useful by connecting them to external platforms and services. For instance, Anthropic recently launched integrations providing applications with access to its AI chatbot Claude.
OpenAI introduced a plugin feature for ChatGPT a few years ago but discontinued it in favor of custom chatbots called GPTs.
"I often hear from users who find ChatGPT’s deep research agent so valuable that they wish it could connect not just to the web but also to their internal resources," said Nate Gonzalez, Head of Business Products at OpenAI, in a blog post on LinkedIn. "[That's why] today we’re launching our first connector."
In addition to answering questions about codebases, the new ChatGPT Deep Research GitHub connector allows ChatGPT users to break down product specifications into technical tasks and dependencies, summarize code structures and patterns, and understand how to implement new APIs using real code examples.
Of course, ChatGPT Deep Research has the potential to generate hallucinations—no existing AI model is immune to confidently fabricating content occasionally. However, OpenAI presents this new feature as a potential time-saving tool rather than a replacement for experts.
An OpenAI spokesperson stated that ChatGPT will respect organizational settings, ensuring users can only see GitHub content they are already allowed to view and codebases explicitly shared with ChatGPT.
OpenAI has been investing in its assisted coding tools, recently releasing an open-source terminal coding tool named Codex CLI and upgrading the ChatGPT desktop application to read code from some developer-focused coding applications. The company views programming as a key use case for its models. For example, OpenAI reportedly reached a $3 billion deal to acquire Windsurf, an AI-powered coding assistant.
In other OpenAI news on Thursday, the company introduced fine-tuning options for developers looking to customize its new models for specific applications. Developers can now fine-tune OpenAI’s o4-mini “reasoning” model using what OpenAI calls reinforcement fine-tuning, a technique employing task-specific scoring to boost model performance. Fine-tuning has also been rolled out for the company’s GPT-4.1 nano model.
According to OpenAI, only verified organizations can fine-tune the o4-mini. Meanwhile, fine-tuning for GPT-4.1 nano is available to all paying developers.
OpenAI began restricting certain models and developer functions after verification, which requires organizations to submit ID and other identification documents, starting in April. The company claims this is necessary to prevent abuse.
