← All leaderboards
Web scraping
What this board tests: fetch one url as clean content/markdown, anti-bot handled. Every provider below ran the identical corpus on 2026-06-20. The composite blends measured quality, latency p50, cost per successful call and error rate. Method and corpora: how we measure.
| Rank | Provider | Composite | Anti-bot bypass | Cost | Latency p50 | Field note |
|---|---|---|---|---|---|---|
| 1 |  Firecrawl |
9.81/10 | 100% | $0.0021 / success | 962 ms | Class-leading result; every anti-bot vendor cleared. |
| 2 |  Scrapfly |
9.48/10 | 100% | $0.0025 / success | 4.5 s | Passes each anti-bot vendor with the single asp=true flag. |
| 3 |  Zyte |
8.63/10 | 100% | $0.00069 / success | 6.5 s | Gets past every anti-bot vendor. |
| 4 |  Bright Data |
7.82/10 | 75% | $0.0019 / success | 3.9 s | |
| 5 |  ScrapingBee |
7.45/10 | 88% | $0.0057 / success | 8.9 s | |
| 6 |  ZenRows |
7.31/10 | 62% | $0.0047 / success | 4.8 s | |
| 7 |  Oxylabs |
7.1/10 | 75% | $0.0013 / success | 10.5 s | |
| 8 |  Crawlbase |
7.08/10 | 75% | $0.006 / success | 3.5 s | |
| 9 |  Spider |
6.73/10 | 50% | $0.00013 / success | 12.1 s | |
| 10 |  ScraperAPI |
6.53/10 | 62% | $0.01 / success | 2.4 s |
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.