PDF rendering endpoint (ptpmcrender.fcgi) — outage or deprecation?

[Wei-Hua Huang]

Hi Europe PMC team,

We use the Europe PMC API as part of our literature retrieval workflow for clinical research projects. I wanted to flag that the PDF rendering endpoint appears to be down, and wanted to check whether this is an outage or an intentional change.

What's broken:

URLs like europepmc.org/articles/PMC5572138?pdf=render and the underlying backend europepmc.org/backend/ptpmcrender.fcgi?accid=PMC5572138&blobtype=pdf are returning empty responses (server closes the TCP connection without sending any data).

What still works fine:

• Search REST API (ebi.ac.uk/europepmc/webservices/rest/search)
• HTML article pages (europepmc.org/article/PMC/...)
• Supplementary files API
• Homepage and all other endpoints
• 
  

When it started:

The endpoint was working normally on March 11, 2026. Downloads began failing sometime between March 11–13. It has been consistently down since — confirmed from both programmatic clients and a standard web browser.

Diagnostics we've run:

• Tested from multiple IPs and with various HTTP clients (curl, Python httpx, Python requests) as well as Chrome browser
• Tested HTTP/1.1 and HTTP/2 — same result
• curl verbose output shows the TLS handshake and HTTP request complete successfully, but the server returns zero bytes
• This appears to be specific to the ptpmcrender.fcgi backend process rather than a network or access issue.

We've been using these URLs as returned by the search API's fullTextUrlList field (where articles are listed with availabilityCode: "F" or "OA"). If this endpoint has been intentionally deprecated, could you point us to the recommended alternative for programmatically retrieving PDFs for articles listed as Free or Open Access in the search API?

Thanks,

Wei Huang

RainCity Analytics

[Mohamed Selim]

Hello Wei

The full text infrastructure has gone through major updates recently, none should affect usual website use cases. Opening pdf on the website still works, but we know many are using  url ?pdf=render.

 It is in progress to provide backward compatible solution for this use case and will update you when it is released.

The fcgi one is decommissioned completely and it was part of intermediary  urls that is not recommended for direct usage.
Regarding bulk download of articles, we encourage you to look at bulk download pages on the website for more information. 

Please note that heavy scrapping of the website for articles causes performance issues on our end and is not recommended way of downloading open access data in bulk.

General info on bulk download datasets available:

https://europepmc.org/downloads

This is the link for openaccess pdf files:

https://europepmc.org/ftp/pdf/

Please let me know if you have any other questions or we can help with anything else.

Kind Regards,

Mohamed