Google has announced Gemini 4 Argon, a new frontier model designed for deep reasoning across complex, long-horizon workflows in software engineering, enterprise knowledge work, and cybersecurity defense. The model is currently being rolled out to trusted cyber defenders through the Fairwind Program while Google engages in a voluntary pre-release process with the U.S. government.
- Argon features an industry-leading 1M output token limit, up from 64K, enabling deeper reasoning in single trajectories.
- It sets new state-of-the-art results on DeepSWE v1.1 (77.9%), Vals Index, AutomationBench (51.3), and LVBench (91.7).
- The model ties for first place on CWE-bench v1 with a score of 68% in vulnerability remediation.
- Pricing is set at $2 per million input tokens and $10 per million output tokens, with cached inputs discounted by 95%.
Google aims to release Argon to developers, enterprises, and consumers soon, starting with paid API customers and Google AI Ultra subscribers, after gathering feedback from early testers to strengthen safety guardrails.