Benchmark · measured in production · Sep 30, 2026
One or a million personalized documents? Our API can take it.
What we printed: a personalized 8-page report with charts, tables and pictures, every copy with its own data, through the same public API you would call, in production.
100 reports in one call
- call 10.12 s
- call 20.06 s
- call 30.05 s
- call 40.12 s
- call 50.08 s
- call 60.03 s
0 s1 s
Two rounds a few minutes apart. Calls 1 and 4 opened each round, after a pause: they are the cold ones.
33–119 ms
of server time per 8-page report, from the warmest call to the coldest.
Many calls at the same instant
Single-document calls, all sent at once. Every call was answered, with no errors or timeouts.
| Calls at once | Answered |
|---|---|
| 400 | 400 / 400 |
| 800 | 800 / 800 |
| 1,500 | 1,500 / 1,500 |
| 2,500 | 2,500 / 2,500 |
That’s our part: printing it.
The times above are server render time, measured on PDFs. The trip to your screen comes on top: your PDF riding fibre optics across the planet. It doesn’t stop for photos. These are the results of this test on Sep 30, 2026, not a promise for every document: a heavier design takes longer.