Gemini 3, the latest AI model from Google, is the company's most intelligent model to date, with more advanced multimodal and vibe coding capabilities, the company said in a blog post on Tuesday. It's available now.
Google says Gemini 3 is "built to grasp depth and nuance" and is better at understanding the intent behind a user's request. The company also touted Gemini 3's multimodal capabilities, such as its ability to turn a long video lecture into interactive flash cards or to analyze a person's pickleball match and find areas for improvement.
Gemini 3 isn't limited to the app. It'll also be available in AI Mode in Search and, for Pro and Ultra subscribers, in AI Overviews. In AI Overviews, Gemini 3 can generate interactive elements.
Don't miss any of our unbiased tech content and lab-based reviews. Add CNET as a preferred Google source.
Google DeepMind's Demis Hassabis and Koray Kavukcuoglu said in a blog post that Gemini 3 Pro is less sycophantic, a problem that's been plaguing AIs and leading to AI psychosis in some. It's also more secure against prompt injection attacks, in which bad actors try to make an AI ignore its original instructions and perform unintended actions.
The company also unveiled Google Antigravity, a new agentic development platform. Google says Antigravity is like having an active partner while making tools or working on projects, autonomously planning and executing complex software tasks while validating its own code. It works in tandem with Gemini 2.5's Computer Use model for browser control and works with nano banana, Gemini 2.5's image model.
Gemini 3's new, more powerful, agentic capabilities will only be available to $250/month Google AI Ultra subscribers at first. This will allow the Gemini Agent to do multi-step workflows, like planning a travel itinerary.
Read more: AI Essentials: 29 Ways You Can Make Gen AI Work for You, According to Our Experts
Google's release of Gemini 3 comes as an AI war is heating up between it, OpenAI, Anthropic and xAI. Google's consistently been leading AI leaderboards, although other AI models haven't been far behind, sometimes trading spots at the top. With the release of Gemini 3, Google seems to be trying to solve for some of AI's more annoying problems, like hallucinations or sycophancy. It's also trying to prove that AIs can be truly agentic, being able to accomplish tasks on the user's behalf. Other agentic models have proven to be problematic in real-world usage and run into various security concerns, especially in web browsers.
The latest AI release from Google also comes at a time when there are fears of an AI bubble forming in the stock market. AI companies, including Nvidia, Google, Meta and Microsoft, account for 30% of the S&P 500. Google currently has a valuation of $3.4 trillion. Even Google CEO Sundar Pichai says the trillion-dollar AI investment boom has "elements of irrationality" and that a burst would affect every AI company, he said in an interview with the BBC. Still, if Google is to keep its stock price moving upward, it needs to demonstrate that its AI models beat the competition.