THE CRUNCH
Google has launched Gemini 4 Argon, its first major frontier model since Gemini 3 in late 2025, after months in which rivals Anthropic and OpenAI appeared to leave it behind. The company claims Argon offers frontier reasoning and improved performance across complex workflows, with less hedging, repetition and cliché-riddled output than typical chatbots, and says it beats GPT-6 Astra, Claude Fable 5.1 and Claude Opus 5.5 on benchmarks spanning coding, cybersecurity, science, finance and law.
Google says it is already using Argon internally, including finding and patching cybersecurity vulnerabilities, halving the computing power needed to run YouTube's comments section, and debugging the drone physics simulator for Alphabet's delivery firm Wing. As with other recent frontier releases, the public rollout follows a limited release to a select group of "trusted cyber defenders" to test for potential risks.
Argon is available today to Pro and Ultra subscribers in the Gemini app and AI mode in Search, and to Ultra subscribers on Google's agentic development platform Antigravity. It is not clear when, or if, it will reach Plus subscribers or free users. That fits a shift Google signalled earlier this month, when it announced that from October 2026 free users would be limited to the smallest Flash-Lite model, a move seen as potentially clearing the way for a more resource-hungry release like Argon.
Pricing is aggressive: an introductory $2 per million input tokens and $10 per million output tokens, roughly a fifth of what OpenAI's GPT-6 Astra and Anthropic's Fable 5 cost. Those prices double when the introductory period ends in late November, but Argon would remain significantly cheaper per token. Google's own caveat applies, though: models can use different numbers of tokens for the same task, so cheaper per-token rates do not guarantee lower real-world costs.