OpenAI introduces Deep Research AI agent for multi research tasks

OpenAI has introduced Deep Research, a new agent that synthesises vast volumes of online data using the o3 model's reasoning capabilities. ChatGPT users can use the Deep Research tool to perform multi-step research jobs. ChatGPT Pro plan members can start using this functionality today with 100 enquiries per month. This new functionality will be available to ChatGPT Plus and Team members in the near future, with ChatGPT Free tier users receiving it later.

To develop a research report, analysts typically spend several hours finding, analysing, and synthesising hundreds of web sources. ChatGPT Deep Research can now complete the identical assignment in just a few minutes. This new capability was trained on real-world jobs that required browser and Python tool use. Deep Research's final report will include precise sources and an explanation of its thoughts. OpenAI promises that this function will be extremely effective at locating niche, non-intuitive information that would otherwise necessitate exploring multiple websites.

Humanity's Last Exam is a novel evaluation that puts AI models to the test on expert-level questions covering a wide range of topics. It includes over 3,000 multiple-choice and short-answer questions covering over 100 topics. This examination saw OpenAI's Deep Research model achieve a record accuracy of 26.6%. OpenAI also claims that this model achieved a new state-of-the-art score in GAIA, another public benchmark for evaluating AI on real-world problems.

Deep Research has the ability to dramatically cut the amount of time spent on research tasks. Its capacity to synthesise material and offer citations could be advantageous for many consumers. Learn more about OpenAI's Deep Research feature here.QuillBot's paraphraser edits your sentences, allowing you to rethink and restate your content quickly and efficiently!