Overview
- Google unveiled Gemini 4 Argon on Wednesday as a flagship model built for long, multi-step work in software engineering, enterprise knowledge tasks and cybersecurity.
- Argon raises the model output limit to an industry-leading 1 million tokens so it can sustain long reasoning trajectories and produce very long results in a single run.
- Access is limited at launch to a cohort of trusted cyber defenders through Google’s Fairwind Program and to internal teams while the company participates in the U.S. voluntary pre-release review and tightens safeguards.
- Google says Argon can autonomously find, validate and patch vulnerabilities and that partner Wiz used it to discover a critical flaw in healthcare software, but most benchmark and use-case claims come from Google and partners and lack independent verification.
- Google provided introductory API pricing of $2 per million input tokens and $10 per million output tokens rising later to $4 and $20, and reported internal wins such as data-center memory savings of 300+ TiB and a roughly 40% improvement on a quantum algorithm that show potential enterprise value.