📷 Image: Times of India (via NewsData)
National
Google launches Gemini 4 Argon AI model, outperforms rivals in key benchmarks
✍️ Times of India
🗓 02 Oct 2026, 05:54 AM
👁 22
Google has unveiled Gemini 4 Argon, its most powerful AI model to date, which outperforms OpenAI's GPT-6 Astra and Anthropic's Claude Opus 5.5 on 13 of 18 benchmarks. The model is aimed at coding, legal, finance and cybersecurity tasks and offers a one‑million‑token output limit.
Sundar Pichai announced the launch of Gemini 4 Argon, describing it as Google's most advanced AI model yet. The unveiling took place as part of the company's ongoing push to lead in generative AI capabilities.
According to Google, Gemini 4 Argon surpassed 13 out of 18 performance benchmarks when tested against OpenAI's GPT-6 Astra and Anthropic's Claude Opus 5.5, showing particular strength in reasoning and task completion.
The model is designed for specialized professional workloads such as software development, legal analysis, financial modeling and cybersecurity defense, and it supports an output limit of up to one million tokens per request.
Initially, access will be granted to cyber defenders through Google's Fairwind Program. Paid API users and subscribers to Google AI Ultra will receive access in the subsequent rollout phases.
According to Google, Gemini 4 Argon surpassed 13 out of 18 performance benchmarks when tested against OpenAI's GPT-6 Astra and Anthropic's Claude Opus 5.5, showing particular strength in reasoning and task completion.
The model is designed for specialized professional workloads such as software development, legal analysis, financial modeling and cybersecurity defense, and it supports an output limit of up to one million tokens per request.
Initially, access will be granted to cyber defenders through Google's Fairwind Program. Paid API users and subscribers to Google AI Ultra will receive access in the subsequent rollout phases.