Google Introduces Gemini 3, Featuring 1M Token Context, Antigravity Automation, and Enhanced Multimodal AI

Google introduced Gemini 3, its most advanced AI model, featuring a 1 million-token context window and PhD-level reasoning. It integrates multimodal intelligence across text, images, video, audio, and code. Powered by the Antigravity Platform, it supports agentic automation and a Deep Think mode for scientific workflows. Gemini 3 outperforms Gemini 2.5 Pro by 50% in reasoning and reliability with enhanced safety measures.