OpenAI’s Deep Research Sets New Standards in AI Capabilities
On Sunday, OpenAI introduced Deep Research, an advanced agentic AI tool designed to tackle complex, multi-step research tasks on the internet. Dubbed as a simulation of a human research analyst, OpenAI claims that what Deep Research accomplishes in ten minutes could take hours for a human to achieve—and it seems the tool is living up to the hype.
Deep Research has already made waves by outperforming other models on one of the most challenging AI benchmarks: Humanity’s Last Exam. Less than two weeks old, the exam features notoriously complex questions crafted by leading experts worldwide. According to benchmarks shared by OpenAI, Deep Research scored an impressive 26.6%, far ahead of competitors like ChatGPT03-mini and DeepSeek’s R1 V3-powered model, which scored 10.5% and 9.4% respectively.
For context, DeepSeek previously held a significant lead with its 9.4% accuracy score. However, OpenAI’s o3-mini model briefly dethroned DeepSeek with a 10.5% score, later boosted to 13% in its high-powered configuration, o3-mini-high. This improvement was attributed to the ability of o3-mini-high to take more time analyzing and reasoning through complex queries. Yet, Deep Research has set a new benchmark, with a 183% increase in result accuracy compared to o3-mini-high.
One of the key reasons for Deep Research’s dominance is its robust search capabilities. The tool scours the web to gather and analyze information for general knowledge questions, giving it a distinct competitive edge. OpenAI’s innovative approach ensures the AI performs at an entirely new level, providing not just answers but well-sourced, in-depth insights.
An OpenAI employee described their experience using the tool as a “personal AGI moment,” stating: “Using Deep Research has been a personal AGI moment for me. It takes 10 minutes to generate accurate and thorough competitive and market research (with sources) that previously used to take me 3 hours.”
Deep Research is proving to be a game-changer, not only redefining the capabilities of AI research tools but also setting the stage for the next evolution of artificial intelligence. For more on Deep Research and other breakthroughs in AI, check out TechRadar’s coverage.
RS News or Research Snipers focuses on technology news with a special focus on mobile technology, tech companies, and the latest trends in the technology industry. RS news has vast experience in covering the latest stories in technology.
1 thought on “OpenAI’s Deep Research Sets New Standards in AI Capabilities”
Comments are closed.