Gemini 4 Carbon takes #1 on Artificial Analysis index, trails Claude Opus 5.5 on coding agents
The benchmarker says Carbon scores 59 but isn't leading everywhere, and isn't public yet.
Artificial Analysis says Google's Gemini 4 Carbon is the new #1 on its Intelligence Index v5 with a score of 59. On its Coding Agent Index it ranks second, behind only Claude Opus 5.5.
The firm credits Carbon's lead to complex reasoning, knowledge and scientific tasks. It says Carbon also made large gains in coding and knowledge work over other Gemini models, but is not leading in all of these areas. As with Gemini 4 Argon, Carbon is being rolled out to selected users and isn't publicly available yet. Google is offering a 50% launch discount with no confirmed end date.
Why it mattersIndependent rankings are where launch claims get tested, and this one shows a split verdict, with Google ahead on overall intelligence and Anthropic still ahead on coding agents.
