Google's strongest model "Gemini 4 Carbon" to be provided sequentially for cyber defense and more
Google announced on the 28th (US time) the next-generation frontier model "Gemini 4 Carbon." It is described as "our strongest model," excellent in cutting-edge cybersecurity and Long Horizon Capabilities.
Google announced the frontier AI model "Gemini 4 Carbon," coming two weeks after the release of "Argon." The company describes it as "our strongest model," stating it excels in cutting-edge cybersecurity and Long Horizon Capabilities. It also features improved resistance to hallucinations and unauthorized manipulation.
The company claims the model achieved the top position in benchmarks, scoring 59 on the Artificial Analysis Intelligence Index and 88.2% on SEC-bench Pro. Regarding its release policy, Google plans to first provide it limitedly to some cyber defense organizations and gradually expand the scope. Access is priced at $2/$10 per million tokens.
This item is a summary. The full, checkable version is at the original source:
Read the original at Impress Watch ↗Read the original at ITmedia NEWS ↗Read the original at Ars Technica ↗Read the original at MarkTechPost ↗