Exactly one month after Google rolled out the heavyweight Gemini 3 Pro flagship AI model, it is now ready to bring Gemini 3’s more efficient ‘Flash’ variant to the masses.
Replacing Gemini 2.5 Flash starting today, Google’s latest offering is now starting to show up for users in the Gemini app and AI Mode in Search.
Designed to bridge the gap between “frontier-level” capabilities and a fast response time, Gemini 3 Flash, despite being built for speed, punches well above its weight class. “With Gemini 3, we introduced frontier performance across complex reasoning, multimodal and vision understanding and agentic and vibe coding tasks. Gemini 3 Flash retains this foundation, combining Gemini 3’s Pro-grade reasoning with Flash-level latency, efficiency and cost,” wrote the tech giant.
The model outperforms the likes of Gemini 3 Pro, Gemini 2.5 Flash, Gemini 2.5 Pro, Claude Sonnet 4.5, GPT-5.2, and Grok 4.1 Fast when it comes to multimodal understanding and reasoning. Similarly, it also outperformed or tied with all the above-mentioned models when it came to multilingual Q&A and long horizon real-world software tasks.
Elsewhere, the model can clearly hold its own with several other benchmarks, including Humanity’s Last Exam, scientific knowledge, math, agentic coding, and more.
The efficient model outperforms Gemini 2.5 Pro in all benchmark tests, all while using 30 percent fewer tokens at $0.50 per 1M input tokens. Gemini 3 Flash is now the default model in the Gemini app.
All of our Gemini users globally will get access to the Gemini 3 experience at no cost, giving their everyday tasks a major upgrade.
Gemini 3 Flash unlocks a faster Search experience
Gemini 3 Flash can do everything that 2.5 Flash did, but better, complete with Gemini 3’s Pro-grade reasoning. This allows the model to analyze short video content and help you come up with plans, analyze audio recordings and identify knowledge gaps, create a custom quiz, give you detailed explanations, and more.
In addition to the Gemini app, Gemini 3 Flash is also starting to roll out as the default model for AI Mode in Search. Google says that the new model should allow AI Mode to better understand nuances in your questions. “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,” wrote the tech giant.
Apart from AI Mode and the Gemini app, Gemini 3 Flash is also available to developers and professionals via the Gemini API in Google AI Studio, Google Antigravity, Vertex AI, Gemini Enterprise, Gemini CLI, and Android Studio.
