Your cache is cold.
Your visitors pay for it.

The first 100 customers get 40% off, forever. Enter the code FOUNDING on the checkout page and the discount applies to every future invoice, plan changes included. The code is limited to 100 redemptions and deactivates itself once the spots are gone.

Preload plugins fill the cache on your server: the page cache, object cache or OPcache. That helps, but your CDN edge cache stays cold everywhere except where a request happens to originate. warmup.rocks warms the CDN layer itself, from many geographic locations, so the copy closest to each visitor is the one that's warm.

Yes, and that's where it shines. Workers Cache (like the classic CDN cache) is per-colo, and it can't be filled from a single warming server. Because our requests genuinely enter Cloudflare's network in different regions, each one warms the colo it lands in.

They get faster too. CDNs like Cloudflare use tiered caching: a cache miss at a small edge location is fetched through a larger upper-tier data center, which caches the response as well. Our warming passes fill those upper tiers, so when a visitor lands on a location we never touched, the request is served by a warm upper tier instead of traveling all the way to your origin.

From dedicated datacenter proxies in 42 countries on 6 continents. Each request enters the CDN network in a different region and warms the edge location it lands in:

Thanks to anycast routing and tiered caching, those entry points fan out into 90+ distinct CDN edge locations per warming pass.

Any CDN that caches at the edge. We auto-detect Cloudflare, Fastly, Amazon CloudFront, Akamai, bunny.net and CDN77 from their response headers and report hit ratio and edge location per provider. You don't need to configure anything.

Warming requests come with an identifiable user agent and can carry a custom header, so you can filter them in your analytics or WAF rules. Server-side analytics can exclude them by header; client-side analytics never sees them at all, since no JavaScript is executed.

Requests are paced and rate-limited, and once the cache is warm most requests are served by the edge and never reach your origin at all. A typical warming run touches your origin roughly once per URL, then the edge takes over.

Nothing. The warmer works without touching your code or your DNS. If your pages are publicly reachable and send cache headers, you're ready. Add your sitemap in the dashboard and start the first run.