Gemini becomes a visual analytics assistant

Most chatbots still answer like a textbook page. Google is now trying to make Gemini answer more like a lab table, with things you can rotate, drag, and test inside the chat itself. (blog.google) A chart is just a picture of numbers, and a simulation is a picture that changes when you move a control. Google said Gemini can now build both in chat, so a question about gravity or sales data can turn into something you can manipulate instead of just read. (blog.google) The new visual layer includes interactive charts, 3D models, and simulations with sliders and editable values. Google’s examples include moon phases, the Earth’s orbit, and data visuals that update when the numbers change. (blog.google) (pcmag.com) Google tied this feature to the Gemini Pro model inside the Gemini app. That means the visual response is not a separate product window or browser tool; it appears in the same conversation where the question starts. (blog.google) (winbuzzer.com) At the same time, Google is pulling NotebookLM closer to Gemini through a new feature called notebooks. NotebookLM is Google’s research tool for asking questions over your own sources, like Portable Document Format files, Google Docs, and websites. (blog.google) (tech.yahoo.com) A notebook in Gemini works like a project folder with memory. Google says you can keep chats, files, and instructions together in one place, then use NotebookLM as the research partner behind that workspace. (blog.google) (techrepublic.com) Google announced notebooks on April 8, 2026, and the interactive visuals on April 9, 2026. Put together, those two updates turn Gemini from a question box into something closer to a working canvas for studying, planning, and analyzing. (blog.google 1) (blog.google 2) This is also Google closing a product gap it helped create. NotebookLM had become one of Google’s clearest artificial intelligence tools for source-grounded research, but it lived off to the side while Gemini stayed the main assistant brand. (blog.google) (dataconomy.com) Now the split is getting smaller: one part of Gemini keeps the documents and context, and another part draws the chart or model that explains the answer. That gives Google a cleaner pitch against rivals that already mix chat with canvases, coding views, or visual workspaces. (blog.google 1) (blog.google 2)