Google announced Gemini 4 Argon on September 30, describing a major step forward for demanding software-engineering, knowledge-work, and cybersecurity-defense tasks. The Chinese post, published a few hours later on October 1, highlights its gains over the earlier Gemini generation.
Performance claims and phased access
Google's launch materials show strong results against leading models, including GPT-6 Astra and Claude Opus 5.5, across several evaluations. Those are benchmark comparisons reported by Google, rather than a promise that Argon will outperform every alternative on every task.
Initial access is limited to trusted cybersecurity defenders through the Fairwind Program and other trusted testers. Google says their feedback will help strengthen safeguards before a wider release, beginning with paid API customers and Google AI Ultra subscribers. The announcement does not mean all subscribers can use Argon immediately.
Read the official Gemini 4 Argon announcement for the rollout details. Separately, our Google AI plan changes guide explains October model-access changes in Gemini Apps and Antigravity.
Adapted from the original Chinese article, published on October 1, 2026.
Comments NOTHING