Google, a division of Alphabet, unveiled the most recent iteration of its artificial intelligence model, Gemini, on Tuesday. The company emphasized that the new features will be instantly accessible in a number of revenue-generating products, such as its search engine.
On paper, Gemini 3, which was released eleven months after the second iteration of the model, looks to maintain Google’s lead in the AI race. Executives emphasized Gemini 3’s top spot on a number of well-known industry leaderboards that gauge the performance of AI models during a press briefing.
In a company blog post, CEO Sundar Pichai called it “our most intelligent model.”
But as Wall Street looks for indications of an AI bubble, the race for AI has shifted more and more from benchmarks to profitable uses of the technology. Alphabet’s cloud computing division’s AI offerings have been a major factor in the company’s financial success thus far this year.
However, despite the support of top developers like Google, OpenAI, and Anthropic, new updates to AI models have struggled to stand out, garnering attention only when they fall short, as was the case with Meta earlier this year.
In contrast to previous releases, Google stressed that Gemini 3 was already supporting a small number of profitable consumer and business products at launch.
During the briefing, Google’s chief AI architect Koray Kavukcuoglu told reporters, “We think Gemini has set quite a new pace in terms of both releasing the models, but getting it to people faster than ever before.”
According to Pichai, Google integrated its new model into its search engine from the very beginning with the launch of the Gemini 3. It used to take weeks or months for new Gemini versions to be incorporated into Google’s most popular products.
Gemini 3 features will be available to paying customers of Google’s premium AI subscription plan in AI Mode, a search function that replaces the web’s standard fare with computer-generated responses for complex queries.
