Save hours of work with Deep Research as your personal research assistant. Now with the ability to draw context from your Gmail, Drive and even Chat in addition to the web, and transform reports into interactive content in Canvas. Try Deep Research today What is Deep Research Get up to speed on just about anything with Deep Research, an agentic feature in Gemini that can automatically browse up to hundreds of websites and even your Gmail, Drive and Chat on your behalf, think through its findings, and create insightful multi-page reports in minutes. With the Gemini 3 model, Deep Research is even better at all stages of research, from planning to delivering even more insightful and detailed reports. Deep Research transforms your prompt into a personalized multi-point research plan Deep Research autonomously searches and deeply browses the web and your Gmail, Drive, and Chat if you choose so, to find relevant, up-to-date information Deep Research shows its thoughts as it reasons over information gathered iteratively and thinks before making its next move Deep Research provides comprehensive custom research reports with more detail and insights, generated in minutes and available as an Audio Overview, saving you hours of time How to use Deep Research Gemini Deep Research is designed to tackle your complex research tasks by breaking them down, exploring sources like across the web, and your Workspace content if you choose it, to find answers, and synthesizing findings into comprehensive results. You can also upload your own files to Deep Research, and make your reports even more immersive by turning them into interactive content, quizzes, Audio Overviews, and more in Canvas. Competitive analysis Build a competitor report that cross-references public web data with your internal strategy memos, feature-comparison spreadsheets, and team chats about a rival product. Investigating a potential sales lead, analyzing a company's products, funding history, team and competitive environment, and merging it with your own notes in Workspace on the client relationship. Topic understanding Diving deep into subjects by comparing and contrasting key concepts, identifying relationships between ideas and explaining underlying principles. Product comparison Evaluating different models of an appliance based on features, performance, price and customer reviews. It’s a step towards more agentic AI that can move beyond simple question-answering to become a true collaborative partner capable of sophisticated thinking and execution. How to access Deep Research Try Deep Research today In 150 countries In 150 countries In 45+ languages In 45+ languages Available to Google Workspace users Available to Google Workspace users

Just select Deep Research in the prompt bar to get started and let Gemini do the research for you. How we built the first Deep Research The day after we pioneered the Deep Research product category on Gemini in December 2024, we gathered some of the team behind the product for a discussion. An agentic system To build Deep Research, we developed a new planning system that enables Gemini app to work through complex problems. For Deep Research, we trained Gemini models to be capable of: Breaking down the problem: When presented with a complex user query, the system first formulates a detailed research plan, breaking the problem into a series of smaller, manageable sub-tasks. You’re in control of the plan: Gemini presents it to you, and you can refine it to make sure it’s focused on the right areas. Breaking down the problem: When presented with a complex user query, the system first formulates a detailed research plan, breaking the problem into a series of smaller, manageable sub-tasks. You’re in control of the plan: Gemini presents it to you, and you can refine it to make sure it’s focused on the right areas. Research : The model oversees the execution of this plan, and intelligently determines which sub-tasks can be tackled simultaneously and which need to be done sequentially. The model can use tools like search and web browsing to fetch information & reason over it. At each step the model reasons over information available to decide its next move. We introduced a thinking panel for users to follow what the model has learnt so far & what it intends to do next. Research : The model oversees the execution of this plan, and intelligently determines which sub-tasks can be tackled simultaneously and which need to be done sequentially. The model can use tools like search and web browsing to fetch information & reason over it. At each step the model reasons over information available to decide its next move. We introduced a thinking panel for users to follow what the model has learnt so far & what it intends to do next. Synthesis: Once the model determines enough information has been gathered, it synthesizes the findings into a comprehensive report. In building the report, Gemini critically evaluates the information, identifies key themes and inconsistencies, and structures the report in a logical and informative way, even performing multiple passes of self-critique to enhance clarity and detail. Synthesis: Once the model determines enough information has been gathered, it synthesizes the findings into a comprehensive report. In building the report, Gemini critically evaluates the information, identifies key themes and inconsistencies, and structures the report in a logical and informative way, even performing multiple passes of self-critique to enhance clarity and detail. New category, new problems, new solutions In building Deep Research, we had to work through three significant technical challenges: Multi-step planning Research tasks require multiple steps of iterative planning. At each step, the model has to ground itself on all information gathered so far, then identify missing information and discrepancies it wants to explore — all while trading off comprehensiveness with compute and user wait time. Training the model to be effective at long multi-step planning in a data efficient manner enabled us to make Deep Research function in an open domain setting across all topics. Long-running inference A typical Deep Research task involves many model calls over several minutes. This creates a challenge for building agents: It has to be built so that a single failure doesn’t mean having to restart the task from the beginning. To address this, we developed a novel asynchronous task manager that maintains a shared state between the planner and task models, allowing for graceful error recovery without restarting the entire task. This system is truly asynchronous: you can hop to a different app or quite literally turn off your computer after starting a Deep Research project and the next time you visit Gemini, you’ll get notified when your research is done. Context management Over the course of a research session, Gemini can process hundreds of pages of content. To maintain continuity and enable follow-up questions, we use Gemini’s industry-leading 1 million token context window complemented with a RAG setup. This effectively allows the system to "remember" everything it has learned during that chat session, making it smarter the longer you interact with it. Evolving with new models When Deep Research launched in December it was powered by Gemini 1.5 Pro. With the introduction of Gemini 2.0 Flash Thinking (experimental) we were able to dramatically improve both the quality and serving efficiency of this product. With thinking models, Gemini takes more time to plan out its approach before it makes its next steps. This innate characteristic of self-reflection and planning makes it a great fit for these kinds of long running agentic tasks. What we see is that now Gemini is even better at all stages of research and delivers more detailed reports. At the same time, the compute-efficiency of the Flash model allows us to expand access to Deep Research to far more users. We’re really excited about developing on flash and thinking models in general and expect deep research to keep getting better and better. And with our most capable model, Gemini 3, Deep Research is even better at all stages of research, delivering even more insightful and detailed reports We built the system to be versatile, so over time we can expand its capabilities by giving you more control over what it can browse and giving it sources beyond the open web. We are excited to see how people use Deep Research, and these real-world experiences will inform how we continue to build and improve Deep Research. Ultimately, our goal is a truly agentic and universally helpful AI assistant. Gemini’s new agentive AI system brings together the best of Gemini, Google Search, and web technologies to continuously search, browse, and think through information in a continuous reasoning loop for more comprehensive results. Check responses. Compatibility and availability vary. 18+.