Two Years On: Every Frontier Model Now Reasons Like o1 Did

ArtificialAnlys · x · 2026-09-26

Artificial Analysis marks two years since o1-preview, the first reasoning model; now every frontier model uses reasoning tokens to think before answering. Its Intelligence Index has grown from 4 exam-style benchmarks (MMLU, GPQA, MATH, HumanEval) to 10 evaluations spanning Terminal-Bench 4.0, SciCode and more, covering 15 of 58 model creators. The site also tracks big tech capex trends and released a 2025 year-end State of AI report.

Related event: Two Years After o1-preview, Reasoning Is Now Standard at the Frontier(2 posts)→

Original post →

More from Infra

Infra channel →