Google I/O 2025: Gemini AI Everywhere, from Search to Smart Glasses
Summary
Gemini receives powerful updates: Gemini 2.5 Pro introduces a parallel “Deep Think” reasoning mode, while Gemini 2.5 Flash receives broader improvements in multimodality, code, and long‐context performance.
Google rolls out tiered AI subscriptions: the previous Premium plan becomes Google AI Pro at $19.99 /month, and a new high‐end Gemini AI Ultra tier costs $250 /month in the U.S., targeting professionals with the most capable models.
Search is transforming into an AI‐first experience, with AI Mode delivering end‐to‐end reasoning for long queries and AI Overviews summarizing results; new labs like “Deep Search” and a ticket‐buying assistant expand research and commerce workflows.
Generative media advances with Veo 3 adding native audio to video generation and Imagen 4 offering 2K resolution text/image synthesis, while the Flow tool lets filmmakers compose scenes, characters, and short clips directly from prompts.
Communication is augmented by Google Beam (formerly Project Starline) for immersive 3D video chat and the Android XR platform for mixed‐reality headsets, including lightweight smart glasses that provide on‐device real‐time translation and navigation.
Core Google services embed Gemini: Gemini Live voice assistant on mobile, smarter Gmail Smart Replies, Chrome page summarization, Meet’s real‐time English/Spanish translation, and Jules an asynchronous AI coding assistant in beta.