On Sunday, OpenAI unveiled a new artificial intelligence (AI) agent within ChatGPT capable of conducting independent research on the Internet. Named Deep Research—distinguishing it from Google’s similarly titled Gemini-powered AI agent—the tool excels at multi-step investigations into complex subjects. OpenAI announced that this new capability is powered by one of the company’s o3 AI models and is exclusively available to ChatGPT Pro users, the highest tier of the subscription service. This launch comes just days after Google introduced its own Deep Research agent for the Gemini Android app.
OpenAI Launches Deep Research AI Agent
In a blog post, OpenAI provided insights into the features of its new AI agent, designed to facilitate research-intensive tasks. The company claims that Deep Research can accomplish in minutes what would typically take a human several hours. This tool complements OpenAI’s existing AI agent, Operator, which can autonomously carry out online tasks such as making reservations, purchasing tickets, and buying goods.
Deep Research in ChatGPT
Photo Credit: OpenAI
OpenAI emphasized the target user base for Deep Research, stating it is particularly beneficial for individuals engaged in detailed research in fields like finance, science, policy, and engineering. Additionally, the tool caters to those who desire highly personalized recommendations for significant purchases that necessitate extensive research, such as cars, furniture, and home appliances.
To access Deep Research, ChatGPT Pro users will find a dedicated button located beneath the text field, adjacent to the Search icon. Users can input their inquiries and click on the icon to activate the AI agent, which will then request further information regarding the user’s preferences and specific requirements. Furthermore, users have the option to upload a file to provide additional context.
Once the user responds to the follow-up questions, the AI agent initiates its research process, with a progress bar indicating advancement. A sidebar will also display a summary of the steps undertaken and the sources consulted. The duration for this output can range from five to thirty minutes, culminating in a detailed report upon completion of the research.
Leveraging a version of the o3 model, the AI agent includes citations when referencing specific sources and employs tables and sub-headings to enhance the report’s readability. OpenAI has indicated that, in the coming weeks, the service will expand to incorporate images, data visualizations, and various analytical outputs in its reports.
At present, Deep Research is limited to ChatGPT Pro subscribers, though OpenAI plans to extend availability to Plus and Teams subscription levels shortly. The ChatGPT Pro subscription is priced at $200 (approximately Rs. 17,500) per month.
