After months sitting on the sidelines of the frontier AI race, Google has eventually launched its long-awaited Gemini 4 Argon model, claiming it can vie alongside the finest the industry has to offer.
Google says Argon delivers “frontier reasoning and improved achievement throughout complex workflows,” during cutting downward on hedging, repetition, and the cliches that continue to plague AI chatbots. The business claims Argon outperforms competitor models akin GPT-6 Astra, Claude Fable 5.1, and Claude Opus 5.5 throughout a range of benchmarks, including coding, cybersecurity, science, finance, and law.
Google says it’s already putting Argon to use internally, specified as finding and patching cybersecurity vulnerabilities, halving the computing power needed to power YouTube’s comments section, and debugging the drone discipline simulator for Alphabet’s drone shipment business Wing.
The example is rolling out today to Google’s higher-tier Pro and Ultra subscribers in the Gemini app and AI manner in Search, as fine to Ultra subscribers on its agentic betterment phase Antigravity. It is not apparent whenever — or if — Argon volition be accessible to Plus subscribers or liberated users.
Google is pricing Argon extremely competitively. The example is launching alongside an introductory cost of $2 per myriad input tokens and $10 per myriad output tokens, about a fifth the cost of OpenAI’s GPT-6 Astra and Anthropic’s Fable 5 models. Even following those expenses twice at the end November — whenever the introductory duration finishes — Argon volition motionless be considerably cheaper than its frontier rivals on a per-token basis, although this doesn’t necessarily translate to lesser expenses in practice, as models can use distinct numbers of tokens for the identical tasks.
The initiate fills in several of the blanks remaining rearward from whenever Google first teased Argon a few weeks ago. In row alongside another latest frontier example releases, the community rollout follows a constricted release to a choose set of “trusted cyber defenders” to be tested for possible risks.
Argon is Google’s archetypal important frontier publish since Gemini 3 in delayed 2025. In the intervening months, notable models from Anthropic and OpenAI have seen Google trail far rearward its rivals. Seismic internal shakeups raised additional doubts concerning the company’s capability — or desire — to reclaim its stance and vie at the frontier of AI.
But Argon’s arrival suggests we can’t figure Google out of the AI competition fair yet. Landing fair a day following Meta announced its own frontier model, Muse Wonder Full, Argon is a reminder of how quickly the balance of power can change in AI, and that the frontier is far from a two-horse race.
Follow topics and authors from this narrative to see additional akin this in your personalized homepage nourish and to obtain email updates.
