Google’s Gemini 4 Argon, released globally today after a limited September 30 rollout to select users, enters the frontier model race with strong emphasis on long-horizon reasoning, coding, and enterprise tasks. On Humanity’s Last Exam (HLE), a 2,500-question expert benchmark designed to resist saturation, current leaderboards show Anthropic’s Claude Fable 5.1 and Opus 5.5 ahead at 59–65 percent accuracy, while Gemini 4 Argon posts 57.1 percent on AA-HLE variants. Traders are watching for official HLE scores from Google, which has highlighted Argon’s edge on Vals Index and DeepSWE engineering benchmarks. Near-term catalysts include developer API adoption, enterprise case studies, and any forthcoming no-tools or tool-augmented HLE evaluations that could shift relative positioning among the top labs.