CALIFORNIA / RankWire.AI / – Google has launched Gemini 4 Argon, its flagship artificial intelligence designed for sophisticated professional applications. Announced on Sept. 30, Argon leads the Gemini 4 series, supporting complex tasks such as software development, financial analysis, legal research, and cybersecurity. It is capable of processing extended reasoning and execution sequences. Currently, access is limited to select cybersecurity experts participating in the Fairwind Program.
The new model raises the maximum output to 1 million tokens, a significant jump from the previous cap of 64,000 tokens. This allows it to handle longer projects without splitting work across multiple sessions. Initial API pricing begins at $2 per million input tokens, while output tokens cost $10 per million. Cached inputs are discounted by 95%. Future pricing is expected to rise to $4 for input and $20 for output tokens.
Many employees within Google already utilize Argon for coding, research, and writing tasks. The model has been tested internally on projects involving data center optimization and large-scale software migrations, such as transitioning C and C++ codebases to Rust and memory profiling across data centers. These efforts have reclaimed over 300 tebibytes of memory, with ongoing analysis revealing additional efficiencies.
Enhanced capabilities for handling extensive technical tasks
On DeepSWE v1.1, Argon scored 77.9%, reflecting its strong performance in extended software engineering assessments. The company has also shared results related to finance, legal work, automation, and multimodal tasks. Developed by Google DeepMind, Argon is part of the broader Gemini model suite, integrating coding tools with long-context reasoning and multimodal functions. Its increased output capacity is designed to support multi-step workflows that require sustained reasoning to complete.
Cybersecurity remains a core focus during the initial deployment. Argon can detect, verify, and patch vulnerabilities in approved defense environments. Through its Scan for Good initiative, Wiz utilizes the model to identify security gaps in public infrastructure. Google reported a 68% score on CWE-bench v1, a benchmark for vulnerability remediation. Selected cybersecurity teams can also use Argon outside standard guardrails for authorized security operations.
Limited public access during phased introduction
Google has not announced a specific date for the wider release of Gemini 4 Argon. The rollout is being conducted in stages, with feedback collected from early users. The company also participates in a voluntary U.S. government process granting pre-release access to cutting-edge AI models. Eventually, access will expand to developers, businesses, and consumers, with paid API users and Google AI Ultra subscribers expected to be prioritized, though no official launch date has been set.
Google confirmed it will not release Gemini 3.5 Pro, which was anticipated prior to Gemini 4. Argon now stands as the flagship for demanding reasoning and professional workloads. Other Gemini variants remain available for different performance and budget requirements. For now, Gemini 4 Argon is limited to trusted testers, cybersecurity partners, and early-access programs, with a broader rollout still pending.