Gemini Deep Research is Google's autonomous AI research agent that transforms complex research tasks into comprehensive, well-sourced reports within minutes. Powered by Gemini 3.1 Pro, the agent autonomously plans multi-step research strategies, searches across the web and user-selected data sources, and synthesizes findings into detailed reports complete with citations and native data visualizations. It is available as a feature within the Gemini app for consumers and as an API for developers. The consumer version integrates deeply with Google Workspace — including Gmail, Drive, and Chat — enabling it to combine public web data with private organizational knowledge. Two variants are available: Deep Research, optimized for speed and interactive streaming ideal for real-time user interfaces, and Deep Research Max, designed for maximum comprehensiveness via extended test-time compute suited for asynchronous background workflows. Deep Research Max achieves 93.3% on DeepSearchQA benchmarks, leading all competitors. Developers can access both agents through the Gemini API using an asynchronous interactions model, with support for MCP server integrations, multimodal inputs (PDFs, images, CSVs), Google Search grounding, code execution, and collaborative planning to refine the research scope before the agent begins. The tool is particularly powerful for market research, competitive analysis, due diligence, literature reviews, and scientific research acceleration, with real-world adoption in financial services, biotech, and enterprise knowledge management.