November 2025 Cloudflare outage

18 November 2025

On 18 November 2025, an outage at Cloudflare, which carries traffic to our API and web app, caused requests to APITemplate.io to fail for up to about two hours. This was not a security incident: Cloudflare has confirmed that the outage was not caused by an attack or any other malicious activity.

What happened

At 11:05 UTC, Cloudflare changed the permissions in one of its database systems. The change made a configuration file used by Cloudflare’s bot protection grow to twice its usual size, beyond a limit in the software that reads it. As the file reached Cloudflare’s network from 11:28 UTC, that software began failing and returning errors instead of passing traffic on to the sites behind it.

The file was rebuilt every five minutes, sometimes correctly and sometimes not, so at first the errors came and went. Eventually every rebuilt copy was faulty, and the errors became continuous. Cloudflare stopped the faulty file from spreading and had most traffic flowing again by 14:30 UTC, and all its services working normally by 17:06 UTC.

How APITemplate.io was affected

Our monitors, which check each service once a minute, recorded these failures on 18 November 2025 (times in UTC):

• API in Singapore, Europe (Frankfurt) and US East (N. Virginia): requests failed at times between 11:34 and 12:49, then failed continuously from 13:00 to 14:35. Each region was unavailable for between 1 hour 40 minutes and 1 hour 55 minutes in total.
• API in Australia (Sydney): requests failed at times between 11:36 and 12:29, for 14 minutes in total.
• Web app (app.apitemplate.io): unavailable from 11:34 to 11:54.

Failed requests received an HTTP 500 error from Cloudflare. Because the errors came and went, some requests may also have failed in the gaps between our checks.

Our own servers were not affected, and no templates, documents or account information were lost or changed.

What you should do

• If your integration does not retry failed requests, a document you requested between 11:28 and 14:40 UTC on 18 November 2025 may not have been generated. If you still need it, send the request again.
• We recommend retrying any request that fails with a 5xx error after a short wait, so that a brief outage does not cost you a document.
• For current availability, see Service Status, where you can also subscribe to updates.

What Cloudflare is doing

Cloudflare has published a full account of the outage. It says it is:

• treating the configuration files its own systems generate with the same caution as input from users;
• adding more ways to switch off a faulty feature across its whole network at once;
• stopping error reports from overwhelming its servers’ resources; and
• reviewing how every core part of its network handles errors.

Questions

Was any of my data exposed?

No. During the outage Cloudflare returned errors instead of passing requests on to us. It did not give anyone access to your data or to our systems.

Who can I contact?

Email support@apitemplate.io. Please don’t include passwords, keys or other secrets in your message.