Try Deep Research, Gemini's new AI research assistant

Deep Research, which is available for Gemini Advanced subscribers, promises to save you valuable time. With it you can now test the optimized version for chat, 2.0 Flash Experimental, on Gemini over the internet.

Recently, Google announced the latest updates for Gemini, including Deep Research, the new agentic mode in Gemini Advanced and the ability to test Gemini 2.0 Flash, its latest experimental model.

Deep Research uses AI to explore complex topics on your behalf and can present its findings in a comprehensive, concise and readable report. Google's new tool offers a first look at how Gemini gets even better at managing complex tasks to save you time.

Deep Research: your personal AI research assistant is here for you

Researching online is not always easy. Imagine you're doing a master's degree and preparing for a robotics exam. You want to learn about trends in autonomous vehicle sensors, understand the different technologies and find out how they are predicted to evolve in the future. Such an assignment might require hours of research and searching through countless open tabs, provided, of course, that you manage to locate that link you surely remember saving somewhere...

Under your supervision, Deep Research takes on the difficult job for you. After you enter your query, it creates a step-by-step research plan that you can edit or approve. Once you approve the plan, the feature begins to deeply analyze relevant information from the web on your behalf.

Within minutes, Gemini repeatedly refines its analysis, browsing the web as you would: it searches, finds interesting information and then starts a new search based on what it has learned. He then repeats this process several times and, when he is done, creates a complete and comprehensive report with the key findings, which you can export in the form of a Google Doc. The report is carefully organized, includes links to the original sources, and puts you in touch with relevant websites, as well as businesses and organizations you might not otherwise find, so you can easily research in depth and learn more. If you want to ask Gemini further questions or improve the report, all you have to do is ask! As a result, you'll have hours of research at your fingertips in just minutes.

Earlier this year Google shared its vision for integrating agentic capabilities into our products and Deep Research is the first feature of the Gemini app that makes this vision a reality. The company developed a new representation system that uses Google's expertise in finding relevant information on the web to guide Gemini's browsing and research. This system, combined with Gemini's advanced reasoning capabilities and industry-leading 1 million token context window, results in the creation of comprehensive and comprehensive reports with useful and readable information.

Deep Research is the ideal tool both for those of you who are new entrepreneurs, starting a small business and want an instant update on the competition, as well as to get suggestions for suitable locations, and for those of you who are marketing professionals looking for recent AI-powered marketing campaigns to use as a benchmark for 2025 planning. The tool is available globally on Gemini Advanced [originally in English] online on mobile and desktop, and is expected to be available on the mobile app soon in 2025. Still, it is also available for Greek users in Greek. Those interested can try it out by selecting "Gemini 1.5 Pro with Deep Research" in the model drop-down and entering your research question.

Gemini 2.0 Flash: Try Google's experimental model

Recently, Google announced Gemini 2.0, the most capable AI model released so far. All Gemini users can now try the optimized version for chat, Gemini 2.0 Flash Experimental, with enhanced performance in a number of key areas as well as speed. With this new model, Gemini 2.0 will unlock an even more useful Gemini assistant. Gemini 2.0 Flash Experimental is accessible via the Gemini model drop-down on the web on desktop and mobile, and will soon be coming to the Gemini mobile app.

Google is excited for you to try it out, however keep in mind that it is an early preview and may not work as expected. Additionally, some Gemini features will not be compatible with this model in its experimental state.

A new era for AI utility

The release of features like Deep Research is a notable step towards Google's goal of developing the world's most useful AI personal assistant. It also represents a step closer to providing representation capabilities through Google products. As Gemini is constantly improving, it can do things on your behalf and increase your productivity.