CALIFORNIA / RankWire.AI / – Google has launched Gemini 4 Argon, its flagship artificial intelligence system designed for advanced professional applications worldwide. Announced on Sept. 30, Argon leads the Gemini 4 series and is capable of supporting intricate software development, financial analysis, legal research, and cybersecurity tasks. It is also adept at managing extensive reasoning and execution sequences. Currently, access remains restricted, with select cybersecurity experts participating via the Fairwind Program.
Argon’s token processing limit has been increased to 1 million, a significant jump from the previous 64,000-token maximum. This enhancement enables the model to handle longer, more complex tasks without dividing work into multiple sessions. API pricing begins at $2 per million input tokens, with output tokens costing $10 per million during initial phases. Cached inputs benefit from a 95% discount. Future prices are expected to be $4 for input tokens and $20 for output tokens.
Numerous employees within Google already utilize Argon for coding, research, and content creation. Internal teams have tested the model on data center optimization, large-scale software migrations, including transitioning C and C++ to Rust, and memory profiling across facilities. These efforts resulted in freeing over 300 tebibytes of memory and discovering further efficiencies through ongoing system analysis.
Enhanced capabilities for extensive technical projects
Google reported a 77.9% score for Argon on DeepSWE v1.1, a benchmark for long-duration software engineering tasks. Additional results cover finance, legal work, automation, and multimodal activities. Developed by Google DeepMind within the broader Gemini model suite, Argon integrates coding tools with long-context reasoning and multimodal functions. Its increased output limit supports complex workflows requiring multiple interconnected steps.
Cybersecurity remains a key focus during the initial rollout. Argon can identify, validate, and patch vulnerabilities within authorized security environments. Through the Scan for Good initiative, Wiz employs the model to detect security issues in public infrastructure. Google also reported a 68% score on CWE-bench v1, a benchmark for vulnerability remediation. Selected cybersecurity teams can utilize Argon outside standard guardrails for approved security tasks.
Limited public access during staged deployment
A specific launch date for widespread public availability of Gemini 4 Argon has not yet been announced. Google is employing a phased release, collecting feedback from early users and participating in a voluntary U.S. government process that grants pre-release access to advanced AI models. Future availability will include developers, businesses, and consumers. Priority access is expected for paid API users and Google AI Ultra subscribers, although no official launch date has been set.
Additionally, the company confirmed it will not release Gemini 3.5 Pro, which was anticipated before Gemini 4. Argon now represents the leading flagship model for sophisticated reasoning and professional workloads. Other Gemini variants remain accessible for users with different needs and budgets. Currently, Gemini 4 Argon is limited to trusted testers, cybersecurity partners, and early-access programs, with broader public rollout still in progress.