← All leaderboards
Document parsing
What this board tests: pdf/document to text, with ocr where the provider supports it. Every provider below ran the identical corpus on 2026-09-14. The composite blends measured quality, latency p50, the track-specific cost metric and error rate. Method and corpora: how we measure.
| Rank | Provider | Composite | Text accuracy | Cost | Latency p50 | Field note |
|---|---|---|---|---|---|---|
| 1 |  Firecrawl |
8.61/10 | 0.91 | $0.0035 / useful | 842 ms | Scanned-PDF OCR sets it apart. |
| 2 |  Jina |
8.05/10 | 0.78 | $0.00013 / useful | 2.1 s | No OCR: scanned PDFs come back empty. |
| 3 |  Spider |
6.37/10 | 0.60 | $3.8e-05 / useful | 8.2 s | Scanned PDFs yield raw bytes. |
Low scores are published alongside high ones — a ranking you can only trust if nothing gets hidden. No per-call markup, no commission from any provider. Cost figures are priced by the 2026-09-14 run's own table; the price the gateway bills today is on each provider page.