Google officially launched Gemini 3, the newest generation of its advanced artificial intelligence model, which the company calls its “smartest and most capable system to date.”

Presented on November 18, the model expands its multimodal capabilities, processing text, images, audio, and video simultaneously while functioning as a digital agent capable of executing complex tasks and generating applications on demand. 

Koray Kavukcuoglu, Google’s Head of AI, emphasized that Gemini 3 represents “a qualitative leap toward deeper, more intuitive and interactive experiences.”

The company stated that the model is trained to capture subtle details across real-world visual and audio contexts, positioning it as its most sophisticated tool for environment-level understanding. 

Gemini 3 is available starting today for users and developers in the United States through the Gemini app, which already reaches more than 650 million monthly users.

Over two billion people will also access its functions through Google Search, where it will be integrated by default. Sundar Pichai, CEO of Google and Alphabet, highlighted that the technology has evolved “from reading text and images to reading the environment” in just two years. 

The launch positions Google in direct competition with OpenAI, following the recent release of GPT-5. The company is providing immediate access to Gemini 3 Pro, while the more advanced Gemini 3 Deep

Think will be rolled out soon for Google AI Ultra subscribers. The model will also be integrated into AI Mode, AI Studio, Vertex AI and the new agentic development platform Google Antigravity. 

As part of its educational initiative, Google announced that all U.S. university students will receive one free year of Google AI Pro. Enhanced coding capabilities within Gemini 3 Pro will also allow for more advanced visualizations and dynamic learning tools.