Search API Benchmark Released: Perplexity Tops the Performance Charts
ArtificialAnlys · x · 2026-08-29
Artificial Analysis has released the Search Index, benchmarking 18 Search API products from 8 providers.
Key Findings:
- Perplexity Search variants (low/medium/high context) occupy the top positions on the leaderboard.
- Cost vs. Performance: Perplexity Search (medium) extends the Pareto frontier, reaching an Index score of 80 at $0.091 per task.
- Context Impact: Increasing search context size reduces the number of searches per task (from 18.0 to 11.4) and lowers total cost, with quality plateauing between medium and high context.
Resources:
- Leaderboard
- Methodology
- Open-source Agent Harness Stirrup
Benchmarks cover DeepSearchQA, BrowseComp, and AA-Omniscience.
Related event: Perplexity Tops Artificial Analysis Search Index(2 posts)→
More from coding & agent
- Prime Agent: A Self-Improving RLM Harness for Coding and Autonomous Tasks — xeophon · 2026-08-29
- Production-grade agent architecture with Hermes and critical telemetry — gregmushen · 2026-08-29
- Interoperability platform supports major AI agent stacks — Daniel_Farinax · 2026-08-29
- Emerging Coding Agent Pattern: Managing Fleets via a Chief of Staff — sergeykarayev · 2026-08-29
- Can Human-LLM Coordination Cut Token Costs Without Changing Weights? — mb3rtheflame · 2026-08-29
- Proofpress Boosts Agent Task Completion, Cuts Unsafe Propagation — tallmetommy · 2026-08-29