Google’s Gemini is gaining the ability to generate interactive 3D models and simulations in response to user questions. As reported by The Verge, the new feature allows Gemini to create interactive 3D models and simulations with controls that let users rotate models, adjust parameters with sliders, and change simulation behavior in real time.
Interactive 3D outputs in the Gemini app
According to The Verge, the feature enables Gemini to produce 3D models and simulations that users can manipulate after the response appears. Instead of static images, the feature provides interaction options such as rotating the AI-generated model, adjusting sliders, and inputting different values that update the simulation in real-time.
The Verge’s hands-on example described a prompt asking Gemini to simulate the Moon orbiting the Earth. In response, Gemini created a 3D model with multiple interaction modes. The interface included a slider to adjust the speed of the Moon’s orbit, a toggle to hide the line showing the Moon’s orbital path, and a button to pause the simulation. The 3D model also supported zooming and rotation, allowing users to examine the geometry from different angles.
For users trying out the feature, the article notes that the experience is built into the Gemini app rather than being a separate tool. Gemini app users can access the capability by selecting the “Pro” model in the prompt bar, then using prompts that request a visualization. After Gemini responds, users can select a “Show me the visualization” button beneath the answer.
How to access the visualization feature
The Verge specifies the interaction pattern. All Gemini app users can access the feature by switching to the “Pro” model in the prompt bar. From there, users can ask Gemini to produce specific visual concepts. The article provides examples of prompts such as “show me a double pendulum” and “help me visualize the Doppler effect.” After Gemini’s response, the user can then select “Show me the visualization.”
This workflow involves a two-step approach: first, the chatbot generates an answer; second, the system offers a visualization entry point tied to the response. The presence of an explicit visualization button indicates that Gemini distinguishes between normal conversational output and a downstream rendering step that produces an interactive 3D object or simulation.
From a technical standpoint, this suggests that Gemini’s “Pro” model interprets visualization requests and maps them into a structured representation that can be rendered with interactive controls like rotation, pausing, and parameter sliders. The Verge’s description of real-time updates when users change values indicates a connection between user inputs and the simulation state within the generated visualization.
Real-time controls and practical applications
In earlier generations of AI image and model generation, users typically received a result that could be viewed but not easily manipulated. According to The Verge, this feature provides controls that allow users to rotate the 3D model, zoom in, pause the simulation, and use sliders to adjust parameters like orbital speed. The article also notes a toggle to hide the orbital path line, which affects what information is displayed during the simulation.
For technical audiences, these interaction elements have practical implications. When a simulation can be adjusted through sliders and updated in real time, it becomes possible to test different scenarios—at least for the types of phenomena Gemini can represent as interactive 3D models. The Verge’s example of adjusting the Moon’s orbit speed demonstrates parameter control, while the “double pendulum” and “Doppler effect” prompts show that Gemini can handle both mechanics-style systems and physics concepts that can be visualized.
The source material does not provide details on limits, accuracy, or how Gemini decides which controls to offer for different topics. The consistency of the visualization interface across different domains and the system’s ability to map user intent into meaningful, interactive controls remain to be observed.
Broader implications for AI-powered interfaces
While the report focuses on Gemini, the broader technology trend is toward chat systems that do more than generate text. The Verge’s description of Gemini producing interactive 3D models and simulations suggests a direction where chat becomes a front end for visual computation—where a user’s question can trigger a rendering workflow with user-adjustable parameters.
In this case, Google’s implementation is tied to specific product details: selecting the “Pro” model, asking for a visualization, and then using the “Show me the visualization” option. This demonstrates how model selection and UI design can be used to provide advanced capabilities within a consumer app.
For the industry, this could influence how competitors design their own multimodal experiences—particularly regarding whether AI-generated visuals are interactive by default and how users transition from a conversational prompt to a manipulable output. The described feature set—rotation, sliders, toggles, pausing, and real-time updates—establishes a baseline expectation for interactivity that other platforms may be measured against.
Source: The Verge