A new name at the top
Independent model evaluation platform Artificial Analysis placed Google's Gemini 4 Carbon model first on its Intelligence Index v5 ranking. The result quickly climbed to the front page of Hacker News.
What the index measures
The intelligence index compares models across multiple task types and condenses the results into a single score. Such independent rankings are one of the references developers and enterprise buyers use when choosing a model.
A follow-up to the launch
Gemini 4 Carbon was announced by Google shortly before; the company said the model delivered notable gains, particularly in coding and agentic tasks. The new ranking is being read as a sign that the claim is backed by outside measurement.
Limited information
The source contains only the ranking headline; details such as score gaps, the other models tested and the evaluation date were not shared. It is therefore not yet clear by how wide a margin the result was achieved.
