The race for leadership in the artificial intelligence sector continues to accelerate with increasingly aggressive strategic moves. After launching the Gemini 3 Pro last month, Google decided to make its AI technology even more universal. The company announced about Global distribution of Gemini 3 Flasha model specifically designed to deliver high responsiveness and cost-effectiveness without compromising the ability to reason. This new version replaces Flash version 2.5 and is the default engine for the free app and search engine AI mode starting today.
The highlight of this release is not only its immediate availability, but also its performance compared to its direct competitors. According to the collected technical data, Gemini 3 Flash not only dramatically improves on the previous generation of Google, but is able to compete with and even surpass GPT-5.2 in certain scenarios.
In multimodal comprehension tests like MMMU-Pro, Google’s variant scored 81.2%, slightly better than the 79.5% recorded by OpenAI’s latest model. Even in such complex assessments as Humanity’s Last Exam, the difference between the two is less than one percentage point, which is a significant fact when you consider that Flash is an optimized and lightweight version. Without a doubt, Google has become a giant that is becoming increasingly difficult for competitors to beat.
Gemini 3 Flash is all about code efficiency and new creative tools
Google has made a special emphasis on the ability to reason and program this model. Gemini 3 Flash showed excellent performance in software development tasks, reaching a score of 78% in SWE-bench Verified, putting it even higher than the Pro version of generation 2.5. Its architecture enables the processing of large amounts of information, such as real-time video analysis or complex data mining, while maintaining very low latency, which is ideal for everyday tasks that require maximum speed.
For the end user, these improvements mean a much faster interaction with multimedia files. Now it is possible upload a video or audio and ask AI to define concepts, create educational quizzes, or create action plans in seconds. In addition, access to Nano Banana Pro, the company’s new image generator, which becomes available from the chatbot itself for users in certain regions, has been integrated, expanding the range of creative tools at no additional cost.
