Google Launches Gemini 4 Argon, Claiming Parity with Rival Frontier Models

Google finally launches its long-awaited Gemini 4 Argon model after months of trailing rivals in the frontier AI race. The company claims Argon delivers "frontier reasoning and improved performance across complex workflows" while reducing the hedging, repetition, and clichés that continue to plague AI chatbots. Google says Argon outperforms competitors like GPT-6 Astra, Claude Fable 5.1, and Claude Opus 5.5 on benchmarks spanning coding, cybersecurity, science, finance, and law.

The company already deploys Argon internally across several projects. The model helps find and patch cybersecurity vulnerabilities, cuts in half the computing power needed to run YouTube's comments section, and debugs the drone physics simulator for Alphabet's drone delivery subsidiary Wing. These internal deployments showcase Google's confidence in the model's real-world capabilities beyond benchmark scores.

Argon rolls out today to higher-tier Pro and Ultra subscribers in the Gemini app and AI mode in Search, as well as to Ultra subscribers on Google's agentic development platform Antigravity. Google does not clarify when or if the model will reach Plus subscribers or free users. Pricing is aggressively competitive, with an introductory rate of $2 per million input tokens and $10 per million output tokens — roughly a fifth the cost of OpenAI's GPT-6 Astra.

Read More at the original source →