Gemini 4 Carbon is the new #1 on the Artificial Analysis Intelligence Index v5
Artificial Analysis placed Google’s Gemini 4 Carbon first on its latest intelligence index, while noting the model was available only to selected users.
TL;DR
- Artificial Analysis said Gemini 4 Carbon scored 59 and ranked first on its Intelligence Index v5.
- The evaluator said the score was driven by complex reasoning, knowledge and scientific tasks; the model ranked first in four of the index’s nine evaluations.
- The model was being rolled out to selected users and was not yet publicly available, limiting independent reproduction of the result.
Artificial Analysis reported that Google’s Gemini 4 Carbon ranked first on the company’s Intelligence Index v5, with a score of 59. The organization said the result was driven by performance in complex reasoning, knowledge and scientific tasks. [1] [2]
The evaluator said Carbon led four of the index’s nine underlying evaluations. It also reported improved coding-agent results but said the model did not lead in coding or knowledge-work categories overall. [1] [2]
Artificial Analysis said Gemini 4 Carbon was being rolled out to selected users and was not yet publicly available. That restricted access means outside users could not readily reproduce the benchmark results at the time of the report. [1] [2]
The headline ranking therefore describes one benchmark suite, not general superiority across all tasks. Wider availability and independent evaluation remain open. [1] [2]
Why it matters
The result adds a new data point to competition among frontier models, but the evaluator’s scope and limited access constrain what the ranking can establish.
Editor's note
Benchmark findings are attributed to Artificial Analysis; model availability was limited when reported.