Google raises AI stakes with Gemini 4 Argon, its most powerful model yet
Google has unveiled Argon, its most powerful AI model yet, as it steps up competition with OpenAI and Anthropic in coding, cybersecurity and other complex workloads
01 October, 2026
TT
16
Google has unveiled Gemini 4 Argon, its most powerful artificial intelligence model to date, as the technology giant steps up its push to compete with OpenAI and Anthropic at the frontier of the rapidly evolving AI market.
The new model anchors Google’s Gemini 4 generation and has been designed to handle complex, long-running tasks spanning software engineering, enterprise work and cybersecurity.
Argon is larger than Google’s previous top-tier Pro models and is being positioned against the most advanced systems from rivals OpenAI and Anthropic, Reuters reported.
Google said Argon was built to “sustain deep reasoning across complex, long-horizon workflows”, with the company initially rolling it out to a select group of cybersecurity organisations through its Fairwind Program.
Broader access will follow in phases as Google tests safeguards and gathers feedback. The company said developers, enterprises and consumers would eventually receive access, beginning with paid API customers and Google AI Ultra subscribers, although it did not give a firm release date.
The model will launch at an introductory price of $2 per million input tokens and $10 per million output tokens, according to Google.

One of Argon’s biggest technical changes is a dramatic expansion in the amount of content it can generate during a single task. Google has increased the model’s output limit from 64,000 tokens to one million tokens, giving it greater capacity to work through lengthy and complicated problems.
Google said thousands of its employees are already using Argon internally for coding, research and writing.
Among the examples highlighted by the company, Argon helped Google researchers improve the efficiency of a quantum computing process by 40 per cent, while teams of AI agents identified data-centre memory optimisations that have freed more than 300 tebibytes of capacity.
Google is also deploying Argon agents to help migrate large C and C++ codebases to the Rust programming language.
Cybersecurity takes centre stage
Cybersecurity is emerging as one of the biggest battlegrounds for increasingly capable AI models.
Google said Argon can autonomously identify, validate and patch critical software vulnerabilities, with trusted cybersecurity partners receiving versions of the model without some of the guardrails that will apply to general users.
Cybersecurity company Wiz is already testing Argon through its Scan for Good initiative. Google said the model identified a critical vulnerability affecting healthcare software used by hospitals that previous frontier AI models had missed.
The company is meanwhile working with the US government under its voluntary process for pre-release access to advanced AI models.
Google’s launch comes as competition among the world’s leading AI developers intensifies. OpenAI and Anthropic have continued to push increasingly capable models into coding and enterprise applications, putting pressure on Google to demonstrate that its Gemini platform can compete at the top end of the market.
Google said Argon delivers performance comparable with frontier models including OpenAI’s Astra and Anthropic’s Opus across key coding and cybersecurity benchmarks.


















