Google releases the powerful Gemini 3 model, achieving a significant leap in inference and coding capabilities.

Google, a subsidiary of Alphabet, has launched its latest version of the artificial intelligence model Gemini 3, emphasizing that these new features will be immediately deployed across multiple profitable products. Executives stated that this version represents a "huge leap" in reasoning and coding capabilities. Like its predecessor, Gemini 3 can process text, images, and other media, and solve complex scientific and mathematical problems. CEO Sundar Pichai said its ability to reason and respond based on input has been significantly improved. Google DeepMind CTO Koray Kavukcuoglu said, "This is our smartest model. It will help people turn any idea they have into reality." Google also introduced Antigravity, a new development platform for building AI-driven coding agents, which is currently open in preview. The system allows developers to delegate tasks to autonomous agents that can write, test, and verify code in editors, terminals, and browsers.