Weeks after the well-received debut of the Gemini 3 series of AI models, Google has announced that it is making Gemini 3 Deep Think mode accessible to all Google AI Ultra subscribers in the Gemini app.
While Gemini 3 comes with multimodal reasoning, the newly rolled out Deep Think mode allows users to leverage the model’s further enhanced reasoning capabilities for tasks involving complex math, science and logic problems, Google said in a blog post on December 4.
It also represents a significant improvement over its predecessor, Gemini 2.5 Deep Think because Gemini 3 Deep Think mode relies on advanced parallel reasoning that lets the AI model explore multiple hypotheses simultaneously.
In terms of benchmarks, Gemini 3 Deep Think scored 41 per cent without the use of tools on Humanity’s Last Exam and 45.1 per cent with code execution on the Hand ARC-AGI-2 benchmark test.
“Gemini 3 Deep Think mode brings our strongest reasoning capabilities yet, now rolling out to @GeminiApp Ultra subscribers. Cool example of how it can simulate complex 3D architecture,” Google CEO Sundar Pichai wrote in a post on X.
The wider rollout of Gemini 3 Deep Think mode comes amid a surge in popularity of Google’s AI offerings, with the tech giant perceived to have edged past OpenAI and taken the lead in the AI race.
Soon after the release of Gemini 3, OpenAI CEO Sam Altman reportedly declared a ‘code red’ situation and told employees in an internal memo that the company was facing “rough vibes” due to “temporary economic headwinds”. More recently, The Information reported that OpenAI was prioritising the development and launch of a new AI model codenamed ‘Garlic’ that is capable of outperforming rival models in coding and reasoning tasks.
Paid users subscribed to Google’s Rs 24,500 per month AI Ultra plan, can try Gemini 3 Deep Think mode by following these steps:
Story continues below this ad
– Open the Gemini app.
– Select ‘Deep Think’ in the prompt bar.
– Select ‘Thinking’ from the model drop down.
– Type your prompt and submit.
Earlier this week, Google announced that it has started testing a new feature that merges AI Overviews with AI Mode, letting users ask follow-up questions or get more information about a topic with a single click.
In a post on X, Robby Stein, the VP of Google Search, said that the integration offers “a new way to seamlessly go deeper in AI Mode directly from the Search results page on mobile.” He added that the merger of AI Mode with AI Overviews will enable users to ask whatever they want to without having to think about how to phrase their questions.
Gemini 3 also marked the first time Google integrated its latest AI models into its search products from the first day of launch.
© IE Online Media Services Pvt Ltd
Leave a Reply