Over the course of a few minutes, Gemini continually refines its analysis, browsing the web like you do: searching, finding interesting information, and then launching a new search based on what it learns. It repeats this process several times and, when finished, generates a full report of the key findings, which you can export to a Google Doc. It’s carefully curated with links to the original sources, connecting you to relevant websites and companies or organizations you might not have otherwise found so you can easily dig deeper to learn more. If you have any follow-up questions for Gemini or want to refine the rapport, just ask! That’s hours of research at your fingertips in just a few minutes.
Earlier this year, we shared our vision to integrate more agentic capabilities into our products; Deep Research is the first Gemini feature to bring this vision to life. We built a new agent system that uses Google’s expertise in finding relevant information on the web to direct Gemini’s navigation and searches. This, combined with the advanced reasoning capabilities of the Gemini model and our 1M token pop-up, creates comprehensive reports with useful, easy-to-read information.
Deep Research is the ideal tool if you are an entrepreneur starting a small business and want to quickly gather competitive analysis and recommendations for suitable locations, or if you are a marketer researching recent marketing campaigns based on AI to compare 2025 planning.
Deep Research is now deployed in Advanced Gemini (from English) on desktop and mobile web, and will be available in the mobile app in early 2025. To get started, toggle the model drop-down menu to “Gemini 1.5 Pro with Deep Research” and enter your question research.