Blind inputs
Provider identities were hidden from the judges, so the inputs carried anonymous codes instead of provider labels.
Search Provider Bench
Which web-search API leads — measured, not marketed.
A fair, blind benchmark showing the top scorer for each search type, what each provider costs, and how much confidence the evidence supports. First look only—no switching recommendation yet.
10providers
216searches
2AI judges
—snapshot date
Results first
Top scorer per search type
Quality score (0-100, 50 = average, higher = better). Scores are relative within each search type and are not absolute grades.
Statistical ties: in every category, the top scorer's 95% confidence range overlaps the runner-up's. These are observed leads, not proven winners.
Search typeTop scorerQuality score
Loading results…
Top-scorer count across all 9 search types · counts do not break the statistical ties above
Calculating…
No switching recommendation yet: this is a first look, and the small gaps should be read with the confidence ranges above.
The local benchmark JSON could not be loaded. Serve this directory as a static site and refresh.
🧾 Extract lane
Top scorer at pulling facts out of a page
How many facts it correctly pulled from the page (higher = better).
Rank · ProviderUsable pages · Median charsFact recallCost per extract
Loading Extract results…
Cost per extract · list price where availableLoading cases…
Dropped providers: loading…
The Extract benchmark JSON could not be loaded. Serve this directory as a static site and refresh.
Price / performance
Quality and API price, side by side.
Compare the Search quality frontier, then scan the published unit price for each provider.
Cost per search vs. average quality
Left is cheaper. Up is better. The strongest value sits toward the top-left.
Provider
Best-value frontier
Calculating best value…
Linkup isn't plotted: no rankable Search categories (insufficient evidence).
API price comparison
What each provider charges
ProviderCost per searchCost per extract
Loading prices…
List prices (vendor pay-as-you-go); real bills vary with volume.
The cost JSON could not be loaded. Serve this directory as a static site and refresh.
Why trust this
Three safeguards, visible in the results.
Two judges
grok-4.5 and gpt-5.6-luna judged independently, so no single judge's quirks decide the outcome.
Confidence intervals
Every score comes with a confidence range; overlapping ranges are reported as statistical ties, not proven wins.
Limits of this release
- Snapshot-only evidence
- Human review of AI judges pending
- No switching recommendation