On October 13, 2026, Google DeepMind globally released Gemini 4 Argon, its newest frontier model, through the AI Studio and Gemini platform APIs, with additional access for Google AI Ultra subscribers in the Antigravity agent platform and for Pro/Ultra subscribers in the Gemini app and Google Search AI Mode. Google claims Argon achieves state-of-the-art performance in programming, law, finance, cybersecurity, and science, ranking first on the Vals Index, which evaluates model performance in economically important knowledge-work domains.
The headline technical upgrade is a 1-million-token output ceiling, allowing Argon to sustain long-form analysis and generation across multi-step workflows such as large-scale code refactoring, legal research, financial modeling, and agentic orchestration. Internally, Google says Argon increased agentic code commits by more than 35% while cutting code rollbacks by nearly one-third. In one case, the Chrome team used Argon to separate a roughly 6-million-line browser-feature legacy folder from Chromium's 40-million-line codebase, completing an architecture refactor across hundreds of files in weeks rather than months.
Commercially, Google is using aggressive introductory pricing: until November 30, 2026, input tokens cost $2 per million and output tokens cost $10 per million, with cached input discounted by 95%; after the promotional window, prices will double to $4 and $20 per million, respectively. That launch price is roughly one-fifth of comparable rates for GPT-6 Astra and Claude Fable 5.1, signaling a direct bid for enterprise and developer share. Google also highlighted Argon-assisted breakthroughs in quantum computing, nuclear fusion, enzyme design, robotics, and cybersecurity patching, positioning the model as a general-purpose accelerator for science and engineering intelligence.
Top comments (0)