Google’s new default AI mannequin: Gemini 3 Flash is quicker and stronger
Abstract created by Sensible Solutions AI
In abstract:
- Google launched Gemini 3 Flash as its new default AI mannequin, providing as much as thrice quicker efficiency than Gemini 2.5 Flash whereas being less expensive.
- PCWorld stories the mannequin excels in multimodal duties, scoring 81.2% in MMMU-Professional benchmarks and performing comparably to Gemini 3 Professional and OpenAI’s GPT-5.2.
- This improve enhances Google’s AI merchandise with improved visible understanding, making superior AI capabilities extra accessible for on a regular basis workflows and information evaluation.
Google has now launched Gemini 3 Flash, a quicker and less expensive AI mannequin based mostly on Gemini 3. Based on Google, Gemini 3 Flash is as much as thrice quicker than Gemini 2.5 Flash, and it outperforms earlier Flash fashions in all inside exams.
In a number of benchmark exams, Gemini 3 Flash carried out on par with each Gemini 3 Professional and OpenAI’s GPT-5.2. Within the multimodal take a look at MMMU-Professional, it even topped the checklist with a results of 81.2 p.c.
The Flash mannequin is meant to be tailored for quick and repetitive workflows. Based on Google, it’s notably appropriate for video evaluation, information extraction, and visible queries. It must also be higher at understanding the person’s intentions and may ship extra visible solutions, for instance with photographs and tables.
Gemini 3 Flash will now be the default mannequin in each the Gemini app and Google’s AI-powered search mode.
This text initially appeared on our sister publication PC för Alla and was translated and localized from Swedish.
