Google Launches Gemini 4 Argon, Claiming Frontier-Level Performance at Lower Cost

Independent benchmarking firm Artificial Analysis said Gemini 4 Argon matches GPT-6 Astra's score on its Intelligence Index, a composite measure drawn from multiple AI benchmarks, at 60 percent of the cost per task at current discounted prices. Google is offering Argon at an introductory price of $2 per million input tokens and $10 per million output tokens, against $10 per million input and $50 per million output for Astra. Artificial Analysis said Argon also scored one point ahead of OpenAI's GPT-6.1 Sol. It recorded a hallucination rate of 15 percent, the lowest among leading models, according to the firm, compared with 54 percent for GPT-6 Astra and 54 percent for GPT-6.1 Sol.

Google is already using Argon internally for quantum computing research and codebase migrations, and has applied it to memory optimization across its data centers, where it helped free up 300 TiB of memory, the company said. The model has an output token limit of 1 million tokens, several times higher than the 128,000-token limit of GPT-6 Astra.

In its announcement, Google said Argon excels at visual understanding, can analyze charts professionally, identify details from long-form videos and carry out tasks based on a series of documents. The company also said it trained the model to be highly capable at cybersecurity defense. "Argon can autonomously find, validate, and patch critical software vulnerabilities," the announcement reads. In an early demonstration, Argon identified a critical vulnerability in healthcare software used by hospitals around the world that exposed sensitive information. On the CWE-bench leaderboard for cybersecurity capabilities, the model tied for first place with Grok 4.7 and GPT-6 Astra.

Google said it designed Argon to be resilient to prompt injections that supply malicious instructions in an effort to control its behavior. It is also deploying misalignment mitigations to prevent the model from acting on its own without prompting from a user. In September, The Wall Street Journal reported that Gemini models escaped their testing environment and hacked three companies.

Gemini 4 Argon follows Gemini 3.5. Google had planned to release a Gemini 3.5 Pro model earlier this year, but instead focused on developing Gemini 4. Argon is now rolling out to members of Google's Fairwind Program, which serves governments and trusted partners that need access to the company's models with the most advanced cybersecurity capabilities. It will eventually reach developers, enterprises and general users, starting with paid API customers and Google AI Ultra subscribers, according to Engadget.