CALIFORNIA / RankWire.AI / – Google has unveiled Gemini 4 Argon, marking its first flagship artificial intelligence tailored for complex professional applications. Announced on Sept. 30, Argon leads the Gemini 4 series, supporting intricate software development, financial analysis, legal research, and cybersecurity tasks. It is also capable of managing extended reasoning and execution sequences. Currently, access remains limited, with select cybersecurity defenders utilizing Argon through the Fairwind Program.
In the recent rollout, Argon increases the maximum token output to 1 million, a significant rise from the former 64,000-token limit. This enhancement enables the model to handle lengthier assignments without splitting work into multiple sessions. Initial API pricing is set at $2 per million input tokens, while output tokens are priced at $10 per million during the same introductory phase. Inputs stored in cache benefit from a 95% discount. Over time, prices are expected to rise to $4 for input and $20 for output tokens.
Already, thousands of employees within Google are leveraging Argon for tasks such as coding, research, and writing projects. Internal teams have also tested the AI on data center optimization efforts and large-scale software migrations. One such project employed Argon agents to facilitate C and C++ migrations to Rust, while another focused on memory profiling across data centers. These optimization initiatives resulted in freeing over 300 tebibytes of memory, with ongoing analyses revealing additional savings.
Argon broadens capabilities for extended technical tasks
As of now, Google reported a 77.9% score for Argon on DeepSWE v1.1, an evaluation metric for long-term software engineering performance. The company also shared results spanning finance, legal processes, automation, and multimodal activities. Developed by Google DeepMind, Argon is part of the larger Gemini model family, integrating coding tools with long-context reasoning and multimodal functionalities. Its expanded output capacity aims to support workflows that involve numerous interconnected steps before completion.
Cybersecurity remains a key focus during this initial phase. Argon can detect, verify, and patch software vulnerabilities within authorized defensive environments. Through its Scan for Good initiative, Wiz utilizes the model to identify security flaws in public infrastructure. Google also reported a 68% score on CWE-bench v1, a benchmark dedicated to vulnerability remediation. Select cybersecurity teams can employ Argon for approved security tasks without standard guardrails, ensuring flexibility during critical operations.
Limited public access continues amid phased deployment
At this stage, Google has not announced a definitive date for widespread public availability of Gemini 4 Argon. The company is deploying a phased rollout, actively gathering feedback from early adopters. It is also participating in a voluntary U.S. government process that grants pre-release access to advanced AI models. Future plans include opening access to developers, enterprise clients, and consumers. Priority is expected to be given to paid API users and Google AI Ultra subscribers, though an exact launch date remains unconfirmed.
Furthermore, Google confirmed that it does not intend to release Gemini 3.5 Pro, which was previously anticipated before the Gemini 4 series. Argon now stands as the flagship model for demanding reasoning and professional workloads. Other Gemini variants remain available to meet different performance and pricing requirements. Presently, Gemini 4 Argon is primarily accessible to trusted testers, cybersecurity partners, and early-access programs, with broader availability still pending.