Perplexity Claims Top Spot on Artificial Analysis Search Index with Dominant Performance

Perplexity has taken the top spot on the Artificial Analysis Search Index, outperforming its competitors by a significant margin. The index, launched on August 18, evaluates search APIs based on their ability to assist AI agents in finding answers. Perplexity's medium Search API configuration achieved an impressive score of 80, surpassing Parallel Search at 75 and Exa Search at 74.

The Artificial Analysis Search Index combines three separate benchmarks: DeepSearchQA, BrowseComp, and AA-Omniscience, totaling 1,700 questions across various research domains. Every search API was evaluated against a fixed GPT-5.6 Luna model running within the Stirrup agent harness framework. This allows for a controlled experiment where only the search tool changes, isolating its value.

The index grades on three dimensions: answer quality, cost per task, and processing time. Perplexity didn't just win overall; its variants dominated the top rankings across different configurations.