Google has made its Gemini 3 Deep Think mode available to all AI Ultra subscribers through the Gemini app. This new feature enhances the model’s reasoning capabilities for complex math, science, and logic tasks, utilizing advanced parallel reasoning to explore multiple hypotheses simultaneously. In benchmark tests, Gemini 3 Deep Think achieved scores of 41% without tools and 45.1% with code execution, outperforming its predecessor, Gemini 2.5. Google CEO Sundar Pichai emphasized the mode’s advanced capabilities, showcasing its ability to simulate complex 3D architecture. This rollout comes as Google gains traction over competitors like OpenAI, which is reportedly focused on developing a new AI model named “Garlic.” Subscribers can access the new mode by selecting ‘Deep Think’ in the Gemini app. Additionally, Google is testing a feature that merges AI Overviews with AI Mode for more interactive search queries. Gemini 3 also marks the integration of AI models into Google’s search products from launch.
Source link
