Perplexity Search tops Artificial Analysis Search Index

ArtificialAnlys · x · 2026-08-29

Perplexity Search API topped the Artificial Analysis Search Index, with its medium context variant scoring 80. Tests reveal Perplexity has the lowest model inference cost per task due to smaller payloads, with total costs around $0.091. The evaluation used GPT-5.6 Luna within a standardized agent framework.

Related event: Perplexity Tops Artificial Analysis Search Index(2 posts)→

Original post →

More from Apps

Apps channel →