Perplexity Search API Sweeps Top Three in Artificial Analysis Search Index

On August 29, benchmarking organization Artificial Analysis published its Search Index, evaluating 18 Search API products from 8 providers. All three Perplexity Search API context variants (low, medium, high) took the top three spots, with the medium configuration scoring 80 and surpassing the previous leader, combining top performance with cost efficiency. The results were published by Artificial Analysis and reshared by Perplexity co-founder Denis Yarats.

Confirmed

• The benchmark covered 18 Search API products from 8 providers; Perplexity's three context variants topped all settings with a clear lead.
• The medium configuration scored 80, beating the previous leader, at roughly $0.091 per task — among the lowest total costs.
• Tests ran in an open-source agent framework with the same model (GPT-5.6 Luna, medium reasoning) for all APIs.
• Perplexity's smaller payload yielded the lowest per-task model inference cost.

Why it matters

• Published by the evaluator itself, the results show Perplexity achieving both performance leadership and cost advantages in the search API space, offering direct guidance for developers choosing search-augmented solutions.

2026-08-29 ~ 2026-08-29 · 6 related posts

• [source] Perplexity Search tops Artificial Analysis Search Index — ArtificialAnlys · 2026-08-29
• [source] Search API Benchmark Released: Perplexity Tops the Performance Charts — ArtificialAnlys · 2026-08-29
• Perplexity Tops Artificial Analysis Search Benchmark Across All Settings — inductionheads · 2026-08-29
• Perplexity Search API Dominates Artificial Analysis Leaderboard with Top Three Spots — camerontstow · 2026-08-29

2 near-duplicate retellings: ChrisUniverse · inductionheads