Google Releases Gemini 4 Argon to Compete With Rivals

Google has released Gemini 4 Argon, a frontier AI model the company says matches rivals including OpenAI's GPT-6 Astra and Anthropic's Claude Fable 5.1 and Claude Opus 5.5, according to The Verge.

The release follows months in which Google trailed competitors in the frontier AI race, The Verge reported. Google says Argon delivers what it calls "frontier reasoning and improved performance across complex workflows," while reducing the hedging, repetition, and clichés the company says continue to affect AI chatbots.

What does Google claim Argon can do?

Google says Argon outperforms rival models across a range of benchmark categories, including coding, cybersecurity, science, finance, and law, per The Verge's reporting. The company has not published a breakdown of specific benchmark scores in the material reviewed for this report.

Google also told The Verge it is already using Argon internally. Cited use cases include finding and patching cybersecurity vulnerabilities, cutting the computing power required to run YouTube's comments section, and debugging the drone physics simulator used by Wing, Alphabet's drone delivery subsidiary. The Verge's account of the YouTube claim describes a reduction in computing power without specifying a percentage figure.

Who gets access to Gemini 4 Argon first?

Argon is rolling out to Google's higher-tier Pro and Ultra subscribers inside the Gemini app and in AI mode within Google Search, The Verge reported. Ultra subscribers also get access on Antigravity, Google's agentic development platform.

The Verge reports it remains unclear whether — or when — Argon will reach Plus subscribers or users on Google's free tier. That staggered rollout mirrors a broader industry pattern of gating the newest models behind paid tiers before wider release.

How is Google pricing the new model?

The Verge reports Google is launching Argon at an introductory price of $2 per million tokens, a figure the outlet describes as competitive within the current market for frontier-model API access. The Verge's published report does not include a complete pricing comparison table against GPT-6 Astra or Claude Opus 5.5 rates.

How does this fit the broader AI subscription market?

Google's tiered rollout of Argon echoes moves by other major platforms to wall off their newest AI capabilities behind premium subscriptions. HTT News previously reported on Meta's push into AI-focused subscription plans, part of an industry-wide shift toward monetizing frontier model access through paid tiers rather than free consumer products.

What is still unknown about Argon's performance?

The Verge's report relies on claims attributed to Google rather than independent third-party benchmark verification. No independently audited benchmark scores, error rates, or head-to-head test results appear in the sourced reporting. Readers evaluating Argon against GPT-6 Astra or Claude Opus 5.5 for specific workflows — coding, legal research, or financial analysis — will need to wait for independent testing data, which has not yet been published in the material reviewed for this report.

Google did not specify in the reporting reviewed when Plus-tier or free users might gain access to Argon.