Google announced Gemini 4 Argon on September 30, 2026, as its new frontier large language model optimized for long-horizon coding, enterprise workflows, and cybersecurity defense, with an industry-leading 1-million-token output limit. It initially rolled out only to internal teams and a small cohort of vetted cyber defenders via the Fairwind Program, with Google emphasizing phased expansion, guardrail iteration, and U.S. government pre-release coordination before broader API and consumer access. Trader sentiment favors an early-October public or paid-API release because executives described making the model available “as soon as possible” after early feedback, creating tight implied probabilities around October 13. The 17% chance of no release by October 31 reflects uncertainty over safety reviews and the absence of any confirmed timeline, while later dates capture the risk of further delays typical for frontier-model launches.