arlen/benchOpen benchmarks for agentic consumers
INDEPENDENT · CC-BY-4.0
UPDATED 14 JUN 2026 · BERKELEY, CA
Vendor Profile · Illustrative prototype

Exa

Exa on the arlen/bench leaderboards — agent web-search accuracy, extraction fidelity, freshness, cost per verified-correct answer, and agent-readiness, scored against golden truth.

Answer firstillustrative

Exa ranks #1 of 4 for agent web-search accuracy at 80.9% hit@5; #1 of 4 for extraction fidelity at 0.74, and is not yet agent-ready in live onboarding trials.

hit@5
80.9%
web search
cost / correct
not yet measured
extraction fidelity
0.74
web extraction
agent-ready
No
onboarding harness
§ A

Web Search — Exa

snapshot web_search-2026-q3
Metric ExaLeaderboard bestDirection
hit@1 %62.2leadshigher is better
hit@5 %80.9leadshigher is better
fresh<30d %higher is better
retrievability hlower is better
cost/correct $lower is better
p50 latency mslower is better
§ B

Web Extraction — Exa

snapshot web_extraction-2026-q2
Metric ExaLeaderboard bestDirection
fidelity 0-10.74leadshigher is better
phrase recall0.67leadshigher is better
boilerplate excl.0.76leadshigher is better
cost/correct $$0.0047leadslower is better
coverage %98.7Jina 99.3higher is better
§ C

Compare Exa

Exa vs SerpAPI Exa vs Brave Exa vs Tavily Exa vs Firecrawl Exa vs Jina Exa vs Serper

Illustrative prototype. No verified vendor run has been published yet; every figure here is a placeholder and must not be cited as a measured result. Numbers are replaced when a snapshot’s first full run lands.