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 6 for agent web-search accuracy at 87.6% hit@5; #5 of 7 for extraction fidelity at 0.74, and is agent-ready — an autonomous agent obtained a working API key in live trials with zero humans.
hit@5
87.6%
web search
cost / correct
$0.0064
web search
extraction fidelity
0.74
web extraction
agent-ready
Yes
onboarding harness
§ A
Web Search — Exa
| Metric | Exa | Leaderboard best | Direction |
|---|---|---|---|
| hit@1 % | 71.2 | leads | higher is better |
| hit@5 % | 87.6 | leads | higher is better |
| fresh<30d % | 81.0 | Tavily 86.4 | higher is better |
| retrievability h | 11.2 | Tavily 7.8 | lower is better |
| cost/correct $ | $0.0064 | Serper $0.0041 | lower is better |
| p50 latency ms | 412 | Brave 301 | lower is better |
§ B
Web Extraction — Exa
| Metric | Exa | Leaderboard best | Direction |
|---|---|---|---|
| fidelity 0-1 | 0.74 | Firecrawl 0.91 | higher is better |
| JS gap Δ | 0.22 | Firecrawl 0.03 | lower is better |
| block rate % | 8.8 | Apify 3.3 | lower is better |
| cost/correct $ | $0.0058 | Jina $0.0009 | lower is better |
| schema validity % | 84.0 | Firecrawl 96.2 | higher is better |
§ C
Exa vs Tavily Exa vs Serper Exa vs Brave Exa vs Firecrawl Exa vs SerpAPICompare Exa
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.