Sundar Pichai introduced Gemini 4 Carbon, Google’s new AI model, as the most powerful system the company has ever developed. Ranked first in version 5.0 of the Artificial Analysis Intelligence Index with a score of 59, it surpassed Claude Opus 5.5 on 58 and GPT-6 Astra on 57. The model also achieved strong cybersecurity results, scoring 88.2% on SEC-bench Pro, 54.8% on ExploitGym and 71.0% on CWE-bench. Its rollout is beginning through the Fairwind Program with a select group of cybersecurity professionals, while broader access and pre-release testing by the US government are planned without a precise public date. Its introductory pricing is $2 per million input tokens and $10 per million output tokens, before a planned increase that will eventually double the rates.
Source: Read the original article
