CALIFORNIA / RankWire.AI / – Google has revealed Gemini 4 Argon, its latest flagship AI designed for sophisticated professional applications. The company announced Argon on Sept. 30, positioning it as the premier model in the Gemini 4 series. This AI is capable of supporting intricate software development, financial analysis, legal research, and cybersecurity tasks. It also manages extended sequences of reasoning and execution. Access to the model remains limited initially, with select cybersecurity teams utilizing Argon through the Fairwind Program.
The maximum output capacity of Argon has been increased to 1 million tokens, a significant jump from the former 64,000-token limit. This enhancement enables the model to handle longer projects without splitting work into multiple sessions. The introductory API pricing begins at $2 per million input tokens, while output tokens cost $10 per million during this period. Cached inputs benefit from a 95% discount. Future prices are set to increase to $4 for input and $20 for output tokens.
Currently, thousands of employees within Google are leveraging Argon for coding, research, and writing tasks. Internal teams have also tested the model on projects such as data center optimization and large-scale software migrations. One initiative employed Argon agents to assist in migrating C and C++ code to Rust. Another focused on memory profiling across data centers, which resulted in freeing more than 300 tebibytes of memory. Teams continue to analyze the same systems, uncovering additional savings along the way.
Argon broadens support for extensive technical tasks
On DeepSWE v1.1, Argon achieved a 77.9% score, reflecting its performance in prolonged software engineering tasks. The company has also published results related to finance, legal work, automation, and multimodal functions. Developed by Google DeepMind as part of the broader Gemini model family, Argon combines advanced coding tools with long-context reasoning and multimodal processing. Its increased output capacity is designed to facilitate workflows that involve multiple interconnected steps before completion.
Cybersecurity remains a key focus during the initial deployment. Argon can detect, verify, and remediate software vulnerabilities within approved defensive environments. Through its Scan for Good initiative, Wiz is utilizing the model to identify security flaws in public infrastructure. Google reported a 68% score on CWE-bench v1, a benchmark focusing on vulnerability remediation. Selected cybersecurity defenders can also operate Argon without standard guardrails when working on approved security tasks.
Limited public access continues amid staged deployment
Google has not yet announced a definitive date for widespread public availability of Gemini 4 Argon. The company is employing a phased rollout approach, collecting feedback from early users. It also participates in a voluntary U.S. government process granting pre-release access to advanced AI models. Eventually, the broader release will include developers, enterprise clients, and consumers. Paid API subscribers and Google AI Ultra members are expected to gain priority access, although no official launch date has been provided.
Google also confirmed that Gemini 3.5 Pro will not be released. This model had been anticipated prior to the Gemini 4 series. Argon now serves as the flagship option for demanding reasoning and professional tasks. Other Gemini models remain available to meet different performance and pricing needs. For the time being, Gemini 4 Argon is primarily accessible to trusted testers, cybersecurity partners, and select early-adopters, with wider availability still in the future.