arlen/benchOpen benchmarks for agentic consumers
INDEPENDENT · CC-BY-4.0
UPDATED 12 JUN 2026 · BERKELEY, CA
Vendor × Web Extraction · Verified

Tavily — Web Extraction

How Tavily performs on the arlen/bench web extraction leaderboard, the verified WCXB run (n=150).

Answer firstweb_extraction-2026-q2

Tavily ranks #3 of 4 on the web extraction leaderboard with fidelity 0.62. Cost per verified-correct is not yet priced (—).

§

Metrics

web_extraction-2026-q2
MetricTavilyvs field
fidelity 0-10.6295% CI 0.58–0.656higher better
phrase recall0.54higher better
boilerplate excl.0.68higher better
cost/correct $lower better
coverage %98.7higher better

Full leaderboard: Web Extraction · snapshot JSON: /bench/api/web_extraction-2026-q2.json

§

Strengths & Weaknesses

by page type

Strongest: article (0.69), service (0.67). Weakest: product (0.51), listing (0.48).

§

Sample Rows

lowest-scoring audited pages
TypePagefidelity
listingAll Latest News0.23
articleTips for Online Success0.45
documentationUsing the Fetch API0.53
§

Vendor Right of Reply

Tavily has not yet been sent its rows for pre-publication review (notifications pending). Right of reply is standing; any response will be published verbatim here and linked from the leaderboard row. No commercial relationship; vendors cannot pay for placement.