Overview
- Google announced Gemini 4 Argon on Wednesday, September 30, 2026, as its new flagship frontier model built for long, multi-step reasoning in software engineering, enterprise work and cybersecurity.
- The model supports an industry-leading one million output tokens, which Google says lets Argon sustain deep, single-run reasoning across very large tasks and documents.
- Argon is being distributed first through Google’s Fairwind Program to trusted cyber defenders and internal teams with an unguarded mode so they can use its full defensive capabilities.
- Google published internal benchmark claims that Argon outperforms many rival frontier models on knowledge and enterprise tests, and said partner Wiz used it to find a critical vulnerability in healthcare software.
- The company set introductory API pricing at $2 per million input tokens and $10 per million output tokens, joined the U.S. voluntary pre-release review, and gave no public date for broader developer or consumer availability.