← 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 2026-06-22. The composite blends measured quality, latency p50, cost per successful call and error rate. Method and corpora: how we measure.
| Rank | Provider | Composite | Text accuracy | Cost | Latency p50 | Field note |
|---|---|---|---|---|---|---|
| 1 |  Firecrawl |
8.82/10 | 0.91 | $0.0011 / useful | 293 ms | Scanned-PDF OCR sets it apart. |
| 2 |  Jina |
8.43/10 | 0.78 | $0 / useful | 263 ms | No OCR: scanned PDFs come back empty. |
| 3 |  Spider |
6.99/10 | 0.60 | $3.8e-05 / useful | 2.0 s | Scanned PDFs yield raw bytes. |
Low scores are published alongside high ones — a ranking you can only trust if nothing gets hidden. The gateway's fee sits on top-ups, not per call, so the order above earns us nothing either way.