Google Unveils Gemini 3 With Major Upgrades in Speed, Safety, and Reasoning
Google has officially launched Gemini 3, the company’s newest and most capable suite of AI models, succeeding Gemini 2.5 introduced earlier this year at Google I/O. The rollout begins today across AI Mode in Google Search, available to Google AI Pro and AI Ultra subscribers, and in the Gemini app, where all users can access the new system.
CEO Sundar Pichai describes the new update as “much better at figuring out the context and intent behind your request, so you get what you need with less prompting,” signaling a major step forward in usability and responsiveness.
Google Gemini 3: Stronger Multimodal Performance and Deeper Interactivity
Google claims that Gemini is now “the best model in the world for multimodal understanding,” capable of delivering richer visualizations and more interactive experiences. Built to analyze and combine text, images, video, audio and even code, the model is designed to help users learn and work across virtually any topic.
The launch starts with Gemini 3 Pro, now available in preview. It is the immediate successor to Gemini 2.5 Pro and, according to Google, “significantly outperforms” it across all major AI benchmarks. The company says Gemini 3 Pro produces answers that are “smart, concise, and direct,” focused on delivering insight rather than generic or overly agreeable responses.
Following the Pro, Google will release Gemini 3 Deep Think, a more powerful but slower model currently undergoing testing by safety reviewers. Once evaluations are complete, it will become available to Google AI Ultra subscribers. Early internal benchmarks suggest Deep Think outperforms the Pro model in reasoning tasks.
Enhanced Security and Safety Testing
Google is also emphasizing security with this release. The company calls Gemini 3 its most secure model to date, noting that it has undergone the most comprehensive safety evaluations ever applied to a Google AI system. Improvements include reduced sycophancy, stronger resistance to prompt-injection attacks, and enhanced protections against misuse through cyberattacks.
With the new version, Google is positioning its AI platform for deeper integration across Search, mobile apps, and premium subscription tiers. As competition intensifies in the rapidly evolving AI landscape, Gemini 3 aims to push Google further into the lead on multimodal capabilities, reasoning power, and overall safety.
Read more of our articles below!
