Artificial Analysis Updates Intelligence Index to v4.2, Reordering Top AI Models

5 Sept 2026, 11:51 am · 7d ago · 1 min read · OfficeChai

Benchmark platform Artificial Analysis has updated its Intelligence Index to version 4.2, adjusting evaluation criteria and reshuffling model rankings. Under the revamped index, Fable 5.1 has captured the top spot, while GPT-6 Astra was ranked second overall. The benchmark adjustment follows community discussions regarding initial test scores where GPT-6 Astra showed limited score differentiation over GPT 5.6 Sol. The updated methodology refines evaluation metrics across complex reasoning, coding, and instruction-following benchmarks to better reflect real-world frontier model capabilities. The ranking shifts underscore intense competition among major AI labs striving for benchmark dominance.