Google is beginning to launch Gemini 3 today, a new series of models the company says is its “most intelligent” and “factually accurate” AI systems yet. They’re also a chance for Google to leap ahead of OpenAI following the rocky launch of GPT-5, potentially putting the company at the forefront of consumer-focused AI models.
For the first time, Google is giving everyone access to its new flagship AI model — Gemini 3 Pro — in the Gemini app on day one. It’s also rolling out Gemini 3 Pro to subscribers inside Search. Tulsee Doshi, Google DeepMind’s senior director and head of product, says the new model will bring the company closer to making information “universally accessible and useful” as its search engine continues to evolve.
“I think the one really big step in that direction is to step out of the paradigm of just text responses and to give you a much richer, more complete view of what you can actually see.”
Gemini 3 Pro is “natively multimodal,” meaning it can process text, images, and audio all at once, rather than handling them separately. As an example, Google says Gemini 3 Pro could be used to translate photos of recipes and then transform them into a cookbook, or it could create interactive flashcards based on a series of video lectures.
You’ll spot some of these improvements across Google’s suite of products, including the Gemini app, where you can build more “full-featured” programs inside the built-in workspace, Canvas. The upgraded AI model will also enable “generative interfaces,” a tool Google is testing in Gemini Labs that allows Gemini 3 Pro to create a visual, magazine-style format with pictures you can browse through, or a dynamic layout with a custom user interface tailored to your prompt.
Gemini 3 Pro in AI Mode — the AI-powered Google Search feature — will similarly present you with visual elements, like images, tables, grids, and simulations based on your query. It’s also capable of performing more searches using an upgraded version of Google’s “query fan-out technique,” which now not only breaks down questions into bits it can search for on your behalf, but is better at understanding intent to help “find new content that it may have previously missed,” according to Google’s announcement.
Google is also not so subtly jabbing at OpenAI, describing Gemini 3 Pro as less prone to the type of empty flattery espoused by ChatGPT. Doshi says you’ll see “noticeable” changes to Gemini 3 Pro’s responses, which Google describes as offering a “smart, concise and direct, trading cliche and flattery for genuine insight — telling you what you need to hear, not just what you want to hear.” The company says it also shows “reduced sycophancy,” an issue OpenAI had to address with ChatGPT earlier this year.
Along with these improvements, Gemini 3 Pro comes with better reasoning and agentic capabilities, allowing it to complete more complex tasks and “reliably plan ahead over longer horizons,” according to Google. The AI model is powering an experimental Gemini Agent feature that can perform tasks on your behalf inside the Gemini app, such as reviewing and organizing emails, or researching and booking travel.
Gemini 3 Pro now sits at the top of LMArena’s leaderboard, a popular platform used for benchmarking AI models. A Deep Think mode enhances the model’s reasoning capabilities even further, though it’s currently only available to safety testers.
Gemini 3 Pro is available inside the Gemini app for everyone starting today, while Google AI Pro and Ultra subscribers in the US can try out Gemini Agent in the Gemini app, along with Gemini 3 Pro inside AI Mode by selecting “Thinking” from the model dropdown.
Read the full article here
