Gemini Deep Research agent just topped the Humanity’s Last Exam (HLE) and DeepSearchQA benchmarks, and now leads BrowseComp—outperforming Google Search and NotebookLM. The results showcase a new AI model’s capabilities and set a fresh standard for open‑source research tools. Curious how it did it? Read the full breakdown. #GeminiDeepResearch #HumanitysLastExam #DeepSearchQA #BrowseComp

🔗 https://aidailypost.com/news/gemini-deep-research-agent-posts-top-results-hle-deepsearchqa-leads