Google Unveils Gemini 4 Argon, a High-Powered Language Model

Google has unveiled its latest language model, Gemini 4 Argon, designed to tackle complex tasks such as finance, software engineering, coding, creative writing, and cybersecurity defense. According to Google, Argon is comparable to other companies' frontier models, like OpenAI's GPT-6 Astra and Anthropic's Opus.

Argon matches GPT-6 Astra's score for its Intelligence Index at 60% of the cost per task, with a hallucination rate of 15%, which is the lowest among leading models. At current discounted prices, Argon costs $2 per million input tokens and $10 per million output.

Google has already used Argon in various applications, including its quantum computing research and codebase migrations, freeing up 300 TiB of memory. The model excels in visual understanding, analyzing charts professionally, identifying details from long-form videos, and performing tasks based on a series of documents.