Google rolls out Gemini 4 Argon first to trusted cyber defenders

The new frontier model starts in a limited rollout while Google tests it with security teams before widening access.

Google wrapped up a busy September AI release cycle with Gemini 4 Argon, a frontier model aimed at difficult reasoning-heavy work and now being tested first with trusted cyber defenders. Argon has a 1-million-token output limit and is being rolled out through the Fairwind Program while Google gathers feedback before opening it more widely. The company also shipped Gemini 3.8 Flash and Gemini 3.8 Flash Cyber, with the former focused on better coding and multi-step work at the same speed and cost as 3.7 Flash, and the latter offering autonomous vulnerability discovery and code patching for trusted organizations. Beyond those models, Google added Gemini 3.8 Live, new text-to-speech tools, Connected Apps in Gemini, and Lyria 3.5, and said the month also included Gemini Notebook upgrades, CC for groups, and broader scientific work such as AlphaGenome Atlas and Project Suncatcher.