← All leaderboards
Screenshots
What this board tests: rendered image of a url. Every provider below ran the identical corpus on 2026-06-21. The composite blends measured quality, latency p50, cost per successful call and error rate. Method and corpora: how we measure.
| Rank | Provider | Composite | Valid image | Cost | Latency p50 | Field note |
|---|---|---|---|---|---|---|
| 1 | 9.5/10 | 1.00 | $0.001 / call | 1.3 s | Highest rate of valid images. | |
| 2 | 9.04/10 | 1.00 | $0.001 / call | 5.9 s | ||
| 3 | 8.57/10 | 0.88 | $0.0028 / call | 2.7 s | ||
| 4 | 8.56/10 | 1.00 | $0.002 / call | 12.0 s | ||
| 5 | 7.78/10 | 0.88 | $0.0049 / call | 27.4 s | ||
| 6 | 7.15/10 | 0.62 | $0.0015 / call | 3.7 s | ||
| 7 | 5.03/10 | 0.50 | $0.001 / call | 8.4 s | ||
| 8 | 2.47/10 | 0.25 | $0.0014 / call | 23.3 s | books.toscrape blocks it — a gap openly noted. |
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.