CALIFORNIA / RankWire.AI / – Google has unveiled Gemini 4 Argon, its premier artificial intelligence model designed for sophisticated professional tasks. Announced on Sept. 30, Argon is positioned as the flagship in the Gemini 4 series, supporting intricate software development, financial analysis, legal research, and cybersecurity functions. The model is also capable of managing extended sequences of reasoning and execution, with initial access currently limited to select cybersecurity experts participating through the Fairwind Program.
The maximum output limit for Argon has been increased to 1 million tokens, representing a significant jump from the previous 64,000-token cap. This expansion enables the model to undertake longer tasks without needing to divide work across multiple sessions. The introductory API pricing starts at $2 per million input tokens, while output tokens are billed at $10 per million during the same introductory period. Cached inputs qualify for a 95% discount. Future pricing is set to rise to $4 for input and $20 for output tokens.
A large number of employees within the company already utilize Argon for coding, research, and writing tasks. Internal teams have experimented with the model on data center optimization efforts and large-scale software migration projects. One such project involved using Argon agents to facilitate C and C++ migrations to Rust. Another focused on memory profiling across data centers. The resulting optimization work liberated over 300 tebibytes of memory, with additional savings identified through ongoing system analysis.
Argon enhances capacity for complex technical tasks
Google reported a 77.9% score for Argon on DeepSWE v1.1, a benchmark assessing extended software engineering performance. Results from its application in finance, legal work, automation, and multimodal tasks were also published. Argon was developed by Google DeepMind as part of the larger Gemini model family, which integrates coding tools with long-context reasoning and multimodal abilities. Its increased output capacity supports projects requiring multiple interconnected steps before completion.
Cybersecurity remains a core aspect of the initial rollout. Argon can detect, verify, and patch software vulnerabilities within approved defensive environments. Through its Scan for Good initiative, Wiz is leveraging the model to identify security flaws in public infrastructure. Google reported a 68% score on CWE-bench v1, a benchmark focused on vulnerability remediation. Certain cybersecurity defenders can also access Argon for approved security tasks without the usual guardrails.
Limited public access expected during gradual deployment
No specific date has been announced by Google for the general public release of Gemini 4 Argon. The company is employing a phased approach, collecting feedback from early users, and participating in a voluntary U.S. government process that grants pre-release access to advanced AI models. Later stages will include availability to developers, enterprise clients, and consumers. Priority access is expected for paid API users and Google AI Ultra subscribers, although an exact launch date remains undecided.
Additionally, the company clarified that it does not intend to release Gemini 3.5 Pro, which was previously anticipated before the Gemini 4 series. Argon now functions as the primary flagship for demanding reasoning and professional workloads. Other models within the Gemini lineup continue to be available to meet different performance and pricing requirements. For the moment, Gemini 4 Argon is mainly accessible to trusted testers, cybersecurity partners, and select early-access programs, with broader availability still pending.