Google announced Gemini 4 Argon, its new frontier large language model, on September 30, 2026, with an immediate limited rollout to vetted cybersecurity defenders via the Fairwind Program. The model leads or ties most internal benchmarks against Claude Opus 5.5 and GPT-6 Astra, particularly on long-horizon software engineering (DeepSWE v1.1 at 77.9%) and enterprise tasks, aided by a 1M-token output limit and stronger multimodal reasoning. Google is following a phased safety approach, including U.S. government pre-release reviews, before expanding to paid API users, Google AI Ultra subscribers, developers, and the public. Broader availability timing, pricing tiers after the introductory $2/$10 per million tokens, and competitive responses from OpenAI and Anthropic remain key swing factors for trader consensus on release windows.