Artificial Analysis overhauls its Intelligence Index after GPT-6 Astra scoring drew skepticism
decoder ·
Artificial Analysis has updated its Intelligence Index to version 4.2, likely in response to skepticism after its earlier scoring put GPT-6 Astra on par with its predecessor while other evaluations showed it leading. The new index gives Astra a four-point gain over its predecessor, but Anthropic's Claude Fable 5.1 still tops the ranking, with Astra second and Meta third. The index adds two benchmarks for real-world knowledge work and PDF analysis, while dropping GPQA-Diamond because models have already solved it. Private test data now makes up 40 percent of the weighting to make gaming the benchmark harder, and the company says it also fixed scoring errors and adjusted grading for stability. Artificial Analysis explained it had delayed updates to keep scores stable during major launches, but fast movement at the top required an interim release. A larger version 5 has been in development for eight months and will be introduced in stages.
decoder