As the competition with Chinese AI platform DeepSeek intensifies, OpenAI has added a new ‘deep research’ mode to ChatGPT. The new feature allows users to conduct a multi-step research for more complex tasks.
ChatGPT’s Deep Research Feature:
today we launch deep research, our next agent.
this is like a superpower; experts on demand!
it can go use the internet, do complex research and reasoning, and give you back a report.
it is really good, and can do tasks that would take hours/days and cost hundreds of dollars.
— Sam Altman (@sama) February 3, 2025
Deep Research marks as OpenAI’s second AI agent, following the launch of Operator AI for browser-related tasks last month and is powered by OpenAI’s latest o3 reasoning model, optimised for web browsing and data analysis. Deep Research searches, interprets and analyses vast amounts of text, images and PDFs on the web to produce a comprehensive report closer to the capabilities of a research analyst, OpenAI claims.
Deep research queries aims to take between 5 and 30 minutes to give results, and the chatbot will send users a notification when their research is complete.
While Deep Research currently only supports text output, OpenAI says it plans to introduce embedded images, data visualisations and other analytical output in the coming weeks.
CEO Sam Altman gave a short instance of how Deep research functions in an X post. He wrote, “people will post lots of great examples, but here is a fun one: i am in japan right now and looking for an old NSX. i spent hours searching unsuccessfully for the perfect one. i was about to give up and deep research just…found it. it is very compute-intensive and slow, but it’s the first ai system that can do such a wide variety of complex, valuable tasks. going live in our pro tier now, with 100 queries per month. plus, team, and enterprise will come soon, and then free tier.”
However, as per the official blog, Deep research, much like other AI tools from OpenAI, has been called prone to ‘hallucination’. OpenAI has noted in its blog that deep research can struggle to differentiate “authoritative information from rumour, and currently shows weakness in confidence calibration, often failing to convey uncertainty accurately”.
Also Read: Apple CEO Applauds Deepseek, Nvidia Launches Deepseek R1 on NIM Platform
