Google unveils Gemini 3, a major AI upgrade with stronger reasoning, multimodal power, and new agentic tools for developers.
Google just fired the next shot in the AI arms race, and it’s a big one. The company has unveiled Gemini 3, calling it its most powerful reasoning and multimodal model yet, and positioning it as the centerpiece of a new era of agentic, deeply interactive AI.
The new release marks a leap in performance across nearly every major benchmark, from mathematics to multimodal analysis.Google says Gemini 3 Pro, rolling out today in preview, is now baked across a suite of its consumer and developer products.“Nearly two years ago we kicked off the Gemini era,” said Sundar Pichai, CEO of Google and Alphabet. “Since then, it’s been incredible to see how much people love it.”Google claims more than 650 million monthly users for the Gemini app and 2 billion monthly users for AI Overviews.With Gemini 3, the company says it is delivering advanced capabilities “faster than ever” due to its full-stack approach to AI.Supercharged reasoning powerGemini 3 Pro outperforms 2.5 Pro on every major AI benchmark. It leads the LM Arena Leaderboard with an Elo score of 1501 and demonstrates PhD-level reasoning with top marks on Humanity’s Last Exam and GPQA Diamond.Beyond text, it pushes multimodal reasoning further with 81% on MMMU-Pro and 87.6% on Video-MMMU. It also delivers a state-of-the-art 72.1% on SimpleQA Verified, a signal of progress on factual reliability.Google says Gemini 3 generates smarter, more concise, more direct responses designed for “genuine insight.”Paired with code-driven high-fidelity visualizations, it can translate complex scientific concepts or create interactive learning tools on the fly.Deeper mode, bigger leapAlongside Gemini 3 Pro, Google is previewing Gemini 3 Deep Think, an enhanced reasoning mode that shows even stronger performance.In testing, it hit 41% on Humanity’s Last Exam, 93.8% on GPQA Diamond, and a record 45.1% on ARC-AGI-2 with code execution.Google says this mode helps the model tackle more novel and complex challenges than ever before.“Every generation of Gemini has built on the last, enabling you to do more,” Pichai said.Gemini 3’s long-context capabilities, multilingual strengths, and multimodal synthesis now support learning workflows like analyzing video lectures, translating handwritten recipes, writing code for interactive flashcards, or even breaking down pickleball match footage.On the developer side, Google is rolling out Google Antigravity, an agent-first development platform built around Gemini 3’s reasoning and tool-use capabilities.The platform gives AI agents direct access to the editor, terminal, and browser for autonomous multi-step execution.Building on this, Gemini 3 tops WebDev Arena, hits 54.2% on Terminal-Bench 2.0, and sets a new high on SWE-bench Verified at 76.2%.Pichai said Gemini 3 “combines all of Gemini’s capabilities together so you can bring any idea to life.” He added that Gemini 3 is “state-of-the-art in reasoning” and better at understanding user intent, making it more intuitive and helpful.Google says Gemini 3 is rolling out today across the Gemini app, AI Mode in Search, AI Studio, Gemini CLI, Vertex AI, and Antigravity. The company expects more models in the Gemini 3 series soon.“We hope you like Gemini 3,” Pichai said. “Much more to come.”
AI Model Deep Think Google Antigravity Google Gemini 3 Google Rollout Multimodal AI Reasoning Benchmark
United States Latest News, United States Headlines
Similar News:You can also read news stories similar to this one that we have collected from other news sources.
Google unveils Gemini's next generation, aiming to turn its search engine into a 'thought partner'Google is unleashing Gemini’s next generation of artificial intelligence in its dominant search engine and other popular online services in the high-stakes battle to create technology that people can trust to enlighten them and manage tedious tasks.
Read more »
Google's new Gemini 3 model arrives in AI Mode and the Gemini appFind the latest technology news and expert tech product reviews. Learn about the latest gadgets and consumer tech products for entertainment, gaming, lifestyle and more.
Read more »
Gemini 3 Is Here—and Google Says It Will Make Search SmarterGemini 3 is skilled at reasoning, generating video, and writing code. Amid talk of an AI bubble, Google notes the new model could help increase search revenue too.
Read more »
Google unveils Gemini’s next generation, aiming to turn its search engine into a ‘thought partner’Google is unleashing its Gemini 3 artificial intelligence model on its dominant search engine and other popular online services in the high-stakes battle to create technology that people can trust …
Read more »
With Gemini 3, Google is flexing its biggest advantage over OpenAIBusiness Insider tells the global tech, finance, stock market, media, economy, lifestyle, real estate, AI and innovative stories you want to know.
Read more »
Google unveils Gemini's next generation, aiming to turn its search engine into a 'thought partner'Google is unleashing its Gemini 3 artificial intelligence model on its search engine in the high-stakes battle to create technology that people can trust to enlighten them and manage tedious tasks.
Read more »