Artificial Analysis releases Intelligence Index v5: Claude Opus 5.5 leads (58), best open model's gap grows from 3 to 12 points
Prepared by Sintelligent with the help of SI. Published in SI: . Updated: .
The new index with private tests shows open models lagging the leaders more than the old version suggested.
Artificial Analysis (post of October 27) rebuilt its index: 60% now consists of evaluations on private sets. Cognition's FrontierCode 2.0 and Terminal-Bench-Science were added; HLE-Diamond, AA-Omniscience v1.1 and CritPt v2 were updated. Claude Opus 5.5 (max) scored 58, GPT-6 Astra (max) 57, GPT-6.1 Sol (max) 55; Anthropic and OpenAI models hold the top five places, followed by Gemini 4 Argon (high) with 50. The best open model, Kimi K3.1 Pro (weights not yet released), scored 46; among models with released weights DeepSeek V4.1 Flash 1017 and GLM-5.3 lead with 32. According to The Decoder, Kimi K3.1 Pro fell from fourth place to ninth and Grok 4.7 lost 12 points. The cost per task among leading models differs by more than 13 times.
Sources 2