Google has unveiled Gemini 4 Argon, its most powerful AI model yet, topping 13 of 18 benchmarks against OpenAI’s GPT-6 Astra and Anthropic’s Claude Opus 5.5. The model targets coding, legal, finance and cybersecurity work, with a 1 million token output limit. It is rolling out first to cyber defenders via the Fairwind Program, with paid API users and Google AI Ultra subscribers next in line.
Related Posts
Meet 10-yr-old Hsu Ting-wei, winner of Silicon Valley invention award
In a remarkable showcase of talent, Taiwanese inventors made waves at the Silicon Valley International Festival. Among them, a ten-year-old…