Across 12 observations, 66 different domains shaped the answers. Only 5 of them appeared in every run of the prompt they turned up on.
MintPDF did not appear in any of the 12 answers. Neither did it appear in the domains the answers drew on. A tool that ran each prompt once would have reported roughly a third of this and shown you a percentage.
Runs in which each brand was named, out of the runs performed. The number in brackets is the order it was mentioned in, which is a proxy for rank and not the model's own ordering.
| Brand | best html to pdf api | html to pdf api for developers | best pdf generation api for s… | alternatives to pdfshift | All prompts |
|---|---|---|---|---|---|
| MintPDFyou | 0/3 | 0/3 | 0/3 | 0/3 | 0/12 |
| DocRaptor | 3/31 · 1 · 1 | 3/32 · 2 · 2 | 3/33 · 1 · 2 | 3/32 · 2 · 2 | 12/12 |
| PDFShift | 3/32 · 2 · 2 | 3/31 · 3 · 1 | 3/32 · 3 · 3 | 3/31 · 1 · 1 | 12/12 |
| Gotenberg | 3/34 · 4 · 4 | 3/34 · 7 · 4 | 3/34 · 4 · 4 | 3/35 · 3 · 3 | 12/12 |
| PDFMonkey | 0/3 | 3/33 · 1 · 3 | 3/31 · 2 · 1 | 3/34 · 5 · 7 | 9/12 |
| Api2Pdf | 2/33 · – · 3 | 1/3– · 6 · – | 0/3 | 3/33 · 6 · 4 | 6/12varies |
| Puppeteer | 2/36 · 5 · – | 1/3– · 5 · – | 0/3 | 3/36 · 4 · 6 | 6/12varies |
| Browserless | 2/35 · 3 · – | 1/3– · 4 · – | 0/3 | 0/3 | 3/12varies |
| PDFCrowd | 0/3 | 0/3 | 0/3 | 2/3– · 7 · 5 | 2/12varies |
4 of 8 brands appeared in some runs of a prompt and not others. That inconsistency is invisible to anything that asks once.
Sources the model retrieved while building its reply. Retrieval is not the same as a visible citation, and we do not report it as one. Where a domain carries competitors and never carries you, that is the gap worth working.
| Domain | Appeared in | Prompts | Consistent | You |
|---|---|---|---|---|
| lightningpdf.dev | 7 of 12 | 3 | varies | not present |
| www.htmltopdfconverter.com.au | 7 of 12 | 4 | varies | not present |
| rendex.dev | 6 of 12 | 2 | every run | not present |
| apiscout.dev | 5 of 12 | 3 | varies | not present |
| docraptor.com | 5 of 12 | 3 | varies | not present |
| hookray.com | 5 of 12 | 2 | varies | not present |
| www.screenshotengine.com | 5 of 12 | 4 | varies | not present |
| apitemplate.io | 4 of 12 | 2 | varies | not present |
| d1030xxn62fyyb.cloudfront.net | 4 of 12 | 3 | varies | not present |
| github.com | 4 of 12 | 3 | varies | not present |
| kiprio.com | 4 of 12 | 3 | varies | not present |
| onesimpleapi.com | 4 of 12 | 2 | varies | not present |
| orshot.com | 4 of 12 | 2 | varies | not present |
| pdfmakerapi.com | 4 of 12 | 2 | varies | not present |
Every number above comes from these. Each one links back to the original answer so you can check us rather than trust us.