Google’s new frontier AI model Gemini 4 Argon goes to cybersecurity defenders first
Google LLC today began rolling out Gemini 4 Argon, a frontier artificial intelligence model that beats rival models from Anthropic PBC and OpenAI Group PBC on most of Google’s benchmarks, to vetted cybersecurity defenders. Outside Google’s own teams, only members of its Fairwind Program can use Argon for now. The program opened Sept. 3 with […] The post Google’s new frontier AI model Gemini 4…
Google has launched Gemini 4 Argon, a frontier AI model designed to secure cybersecurity defenses ahead of wider public release. Among the select group of cybersecurity defenders who can currently access Argon are CrowdStrike and Palo Alto Networks. The model outperforms Anthropic's Claude Opus 5.5 and OpenAI's GPT-6 Astra on Google's benchmarks, scoring 77.9% on long-horizon software engineering tasks.
Google has engineered Argon to excel in complex, multi-step work, limiting its output to 1 million tokens to match. Thousands of Google staff are already leveraging Argon for large-scale engineering projects, including porting C and C++ code to Rust and optimizing video decoding in their Fuchsia operating system. An internal model with Argon's protections removed will also hunt for software vulnerabilities, fixing CWE-bench v1 issues as well as a critical healthcare software flaw discovered by Wiz Inc.'s Scan for Good program.
Google plans to charge $2 per million input tokens and $10 per million output tokens upon launch, with paid developers and Google AI Ultra subscribers joining later.
Written by urgent.news from SiliconANGLE's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
Also reported by 9 other outlets
- Google launches Gemini 4 Argon to reclaim the AI frontier businessinsider.com
- Google unveils long-awaited Gemini 4 axios.com
- Google announces Gemini 4 and says it’s so capable that only ‘trusted cyber defenders’ can have it right now theverge.com
- Google announces Gemini 4 flagship AI model after months of delays freemalaysiatoday.com
- Sources: some Google employees say Gemini 4 performs well on benchmarks but struggles with some real-world coding tasks; Google disputes that characterization (Bloomberg) bloomberg.com
- Google rolls out Gemini 4 Argon, its most advanced AI model cnbc.com
- Google releases most advanced Gemini AI model ft.com
- Google announces Gemini 4 flagship AI model after months of delays businesstimes.com.sg
- and 1 more
