V4.1 ranks fifth on LiveBench, first in agentic coding, a user reports
The user credits Python performance for the coding lead, while reporting a TypeScript tie with Fable 5.1 and Muse Spark 1.3 at 60%.
TLDR
A user discussing LiveBench results on September 11 places V4.1 fifth overall and first in the Agentic Coding category. They attribute the coding lead to Python performance. Their language-by-language breakdown puts V4.1 level with Fable 5.1 and Muse Spark 1.3 in TypeScript at 60%, with a JavaScript score of 81.8 versus 77.3 for each of the next four models.
Combined views
21.5K
4 Sources, first seen 19d ago