Google launches Gemini 3: Now live on Gemini app, Search
What's the story
Google has finally launched Gemini 3, touted to be its most sophisticated, most intelligent, and factually accurate artificial intelligence (AI) model yet. In a blog post, Alphabet CEO Sundar Pichai even referred to Gemini 3 as the tech giant's most powerful agentic and vibe-coding model to date. Moreover, the LLM is now accessible through the Gemini app, AI Mode in Google Search, Google AI Studio, Vertex AI, and Gemini CLI.
Twitter Post
Pichai unveils highlights of Gemini 3 rollout
Introducing Gemini 3 ✨
— Sundar Pichai (@sundarpichai) November 18, 2025
It’s the best model in the world for multimodal understanding, and our most powerful agentic + vibe coding model yet. Gemini 3 can bring any idea to life, quickly grasping context and intent so you can get what you need with less prompting.
Find Gemini… pic.twitter.com/JI7xKkAZXZ
Model introduction
Gemini 3 Pro: Google's new flagship AI model
Google is also making its new flagship AI model, Gemini 3 Pro, available to everyone from day one, rolling out this model to subscribers within the Search platform. Tulsee Doshi, senior director and head of product at Google DeepMind, noted that this advanced model will bring the company closer to its goal of making information "universally accessible and useful."
Advanced features
Gemini 3 Pro's multimodal capabilities
Interestingly, Gemini 3 Pro is "natively multimodal," meaning it can process text, images, and audio simultaneously. This capability could be quite useful for tasks such as translating photos of recipes into a cookbook or creating interactive flashcards from a series of video lectures. According to Google, the new model will also enable "generative interfaces," a tool currently being tested in Gemini Labs that lets users create magazine-style visual layouts with pictures they can browse through.
Product integration
Gemini 3 Pro's impact on Google products
The improvements brought by Gemini 3 Pro will be seen across Google's product suite, including the Gemini app, said the company. The upgraded AI model will also power an experimental feature called Gemini Agent. It can perform tasks on behalf of users within the Gemini app, such as reviewing and organizing emails or researching and booking travel plans. Meanwhile, there's also a Deep Think mode that further enhances the model's reasoning capabilities, but it is currently available to safety testers.
Market position
Gemini 3 Pro tops LMArena leaderboard
Notably, Gemini 3 Pro now leads the LMArena leaderboard, a popular benchmarking platform for AI models. It is available in the Gemini app for all users, while Google AI Pro and Ultra subscribers in the US can test out Gemini Agent within the app as well.