On July 29, 2026, Alphabet Inc. launched coordinated updates across its AI portfolios, including Lyria 3.5, a music generation model within Google Flow Music, upgrades for the Gemini macOS application, and features for YouTube advertisers. Lyria 3.5, developed by Google DeepMind, is available immediately via the web-based Google Flow Music studio and allows users to generate fully arranged, vocal-accompanied tracks up to three minutes in length from a text prompt. It features improvements in structural awareness and prompt adherence, generating realistic, emotionally nuanced vocals. New features include Selective Section Painting for editing specific stems, Granular Control over BPM, drum intensity, and track duration, and accessibility within the Chrome browser. For Gemini on macOS, users can activate a voice dictation protocol by long-pressing the Fn key, which operates inside any active application window, automatically editing out filler words and handling mid-sentence corrections.