Stories
Google Unveils Flagship AI Model "Argon" After Delays, Aims to Rival OpenAI and Anthropic
- •Google announced "Argon," a new top-tier AI model for its Gemini 4 generation, aiming to compete with OpenAI's Astra and Anthropic's Opus.
- •Argon is larger than Google's previous "Pro" models and is described as its most performant model for complex workloads, comparable to frontier models on coding and cyber benchmarks.
- •The model is currently being provided to select cybersecurity partners, with no public release timeline announced.
- •Argon reportedly outperformed Astra and Opus on some industry benchmarks but lagged on others, including two coding-related benchmarks.
- •Google has canceled the release of Gemini 3.5 Pro and has undergone an overhaul of its DeepMind AI lab amidst delays in its AI model development.
Key Questions Answered
Google's new flagship model, Argon, is larger than previous Pro models and built for complex workloads. It is comparable to frontier models like OpenAI's Astra and Anthropic's Opus on key coding and cyber benchmarks, though it lagged on a few coding metrics.
Google no longer plans to release Gemini 3.5 Pro, which CEO Sundar Pichai originally scheduled for a June launch. Amid months of delays, the DeepMind AI lab was overhauled, with founder and CEO Demis Hassabis stepping aside and several Gemini model leaders leaving the company.
Google is providing the Argon model to select cybersecurity partners and participating in the Trump administration's voluntary process for pre-release model access, though it has not shared a timeline for a public release.
