Google has launched an advanced feature called Deep Think for Ultra subscribers of its Gemini app, powered by a high-level AI model that previously achieved Bronze-level performance in the International Mathematical Olympiad (IMO). This upgrade facilitates advanced reasoning, scientific problem-solving, and coding assistance through innovative parallel thinking techniques.
Deep Think enables users to tackle complex problems rapidly while maintaining multi-step reasoning capabilities, making it invaluable for coders, researchers, and creators. Users can access this feature by opening the Gemini app, selecting the 2.5 Pro model, and toggling on Deep Think in settings. Although daily usage is initially limited, Google plans to expand access through the Gemini API. CEO Sundar Pichai has praised the upgrade, highlighting improved safety and reliability. This major enhancement marks Google’s commitment to making expert-level AI tools accessible to everyday users, reinforcing its leadership in AI technology.
Source link
