Google launches Gemini 3 Flash for frontier performance on PhD-level benchmarks

Google announced the launch of its high-speed, low cost Gemini 3 Flash AI model that the company says can deliver Gemini 3’s Pro-grade reasoning along with low latency and high efficiency, at a far lesser cost while also enabling agentic workflows.

The launch of Gemini 3 Flash comes after Google released its Gemini 3 Pro and Gemini 3 Deep Think mode last month. These models were well-received by users, with media outlets reporting that OpenAI CEO Sam Altman had declared a “code red” at the ChatGPT-maker in order to improve their own models as soon as possible.

The search and internet giant claimed that Gemini 3 Flash could deliver frontier performance on PhD-level reasoning and knowledge benchmarks. Gemini 3 Flash also outperformed Google’s best 2.5 model, Gemini 2.5 Pro, across a number of benchmarks, per the company. According to its comparison, Gemini 3 Flash also beat out Claude Sonnet 4.5 and GPT-5.2 across several benchmarks.

“Building on the reasoning capabilities of Gemini 3 Pro, AI Mode with Gemini 3 Flash is more powerful at parsing the nuances of your question. It considers each aspect of your query to serve thoughtful, comprehensive responses that are visually digestible — pulling real-time local information and helpful links from across the web. The result effectively combines research with immediate action: you get an intelligently organized breakdown alongside specific recommendations — at the speed of Search,” noted Google in its blog post.

“This shines when tackling complex goals with multiple considerations like trying to plan a last-minute trip or learning complex educational concepts quickly,” it added.

Google announced that Gemini 3 Flash would be made accessible to everyone across Google products, via offerings such as Google AI Studio, Gemini CLI, Google Antigravity, the Gemini app, AI Mode in Search, Vertex AI, and Gemini Enterprise.