Follow the latest coverage, related explainers and connected technology stories.
Google reviewed a set of artificial intelligence updates it announced during September 2026, led by the Gemini 4 Argon model with a context window of one million tokens and the Fairwind program for testing its security capabilities. The updates included audio models, new integrations with applications, tools for Windows, and research projects for mapping the genome and tracking methane from space.
Google has announced the availability of Gemini 3.8 Live, Gemini 3.8 Live Extended Thinking, and Gemini 3.5 Transcribe for developers through the Gemini API and Google AI Studio. The models are designed to build voice agents capable of maintaining conversations and completing tasks, alongside low-error speech-to-text conversion across more than 85 languages.