Google rolls out Gemini 4 Argon, its most advanced AI model
Google's Gemini 4 Argon claims the top spot in coding and cybersecurity. Here's what changed under the hood — and what it means for the frontier race.
Why it matters
Google ships its most capable model yet with claimed improvements in coding, cybersecurity, and professional reasoning. This is a frontier capability release that re-enters Google into the active model-race conversation after months of incremental updates. Practitioners need to know: GA status, pricing, context window, performance deltas vs Claude/GPT-6, and whether the cybersecurity claims are independently validated or vendor benchmarks.
The key facts
8 to knowModel name: Gemini 4 Argon
Release date: September 30, 2026
Claimed improvements: coding, cybersecurity, complex professional work
Status: released (CNBC reporting; no explicit GA confirmation in headline)
Benchmark data: not disclosed in headline or snippet
Pricing/availability: not disclosed
Context window: not disclosed
Comparison to Claude, GPT-6: not stated
The story so far
Earlier coverage of this storyline
- Google plans Gemini 4 release before year-endComputerworld
- This story
Go to the source
CNBC Technologycnbc.com
Publisher excerpt: Gemini 4 Argon is Alphabet's most advanced model yet, with major coding, cybersecurity, and complex professional work improvements.