Artificial Analysis Intelligence Index v5: Kimi K3.1 Pro leads open weights at 46, DeepSeek V4.1 Flash at 32
The post details v5 changes: 60% private datasets to cut contamination plus new FrontierCode 2.0 and Terminal-Bench-Science tests, with Kimi K3.1 Pro at 46 and DeepSeek V4.1 Flash tied with GLM-5.3 at 32 among open weights.
Já acompanho de perto os benchmarks da @ArtificialAnlys para tomada de decisão na escolha de modelos abertos na empresa, e o novo Intelligence Index v5 subiu bastante a régua: • 60% do índice agora roda sobre datasets privados (cortando contaminação de benchmark) • Novos testes práticos de código e terminal (FrontierCode 2.0 e Terminal-Bench-Science) • Nos open-weights, Kimi K3.1 Pro (46 pts), DeepSeek V4.1 Flash e GLM-5.3 (32 pts) puxando a fila
This post quotes another post: x.com/artificialanlys/status/2115265133467635877