After months sitting on the sidelines of the frontier AI race, Google has finally launched its long-awaited Gemini 4 Argon model, claiming it can compete with the best the industry has to offer.<\/p>\n
Google says<\/a> Argon delivers \u201cfrontier reasoning and improved performance across complex workflows,\u201d while cutting down on hedging, repetition, and the cliches that continue to plague AI chatbots. The company claims<\/a> Argon outperforms rival models like GPT-6 Astra, Claude Fable 5.1, and Claude Opus 5.5 across a range of benchmarks, including coding, cybersecurity, science, finance, and law.<\/p>\n Google says it\u2019s already putting Argon to use internally, such as finding and patching cybersecurity vulnerabilities, halving the computing power needed to power YouTube\u2019s comments section, and debugging the drone physics simulator for Alphabet\u2019s drone delivery company Wing.<\/p>\n The model is rolling out today to Google\u2019s higher-tier Pro and Ultra subscribers in the Gemini app and AI mode in Search, as well to Ultra subscribers on its agentic development platform Antigravity. It is not clear when \u2014 or if \u2014 Argon will be available to Plus subscribers or free users.<\/p>\n Google is pricing Argon very competitively. The model is launching with an introductory price of $2 per million input tokens and $10 per million output tokens, around a fifth the cost<\/a> of OpenAI\u2019s GPT-6 Astra and Anthropic\u2019s Fable 5 models. Even after those costs double at the end November \u2014 when the introductory period finishes \u2014 Argon will still be significantly cheaper than its frontier rivals on a per-token basis, though this doesn\u2019t necessarily translate to lower costs in practice, as models can use different numbers of tokens for the same tasks.<\/p>\n The launch fills in some of the blanks left behind from when Google first teased Argon<\/a> a few weeks ago. In line with other recent frontier model releases, the public rollout follows a limited release<\/a> to a select set of \u201ctrusted cyber defenders\u201d to be tested for potential risks.<\/p>\n Argon is Google\u2019s first major frontier release since Gemini 3 in late 2025. In the intervening months, impressive models from Anthropic and OpenAI have seen Google trail far behind its rivals. Seismic internal shakeups<\/a> raised further doubts about the company\u2019s ability \u2014 or desire<\/a> \u2014 to reclaim its position and compete at the frontier of AI.<\/p>\n