CDN Cache Hit Ratio & Origin Offload Calculator

Understanding the CDN cache hit ratio and origin offload calculator — concept and principle

A content delivery network (CDN) keeps copies of your files on edge servers close to users. Each request is either a hit, served from the edge cache, or a miss, which has to be fetched from your origin server. The cache hit ratio measures how well the CDN shields the origin, and it drives both origin bandwidth costs and page speed. This calculator gives the hit ratio and an estimate of how much traffic still reaches the origin.

With a 95% hit ratio only one request in twenty reaches the origin — 500 GB instead of 10,000 GB.

How it is calculated

Hit ratio (%) = hits ÷ (hits + misses) × 100.
Origin traffic (GB) = total traffic × misses ÷ (hits + misses).
Offload is the same as the hit ratio: the share of requests the origin never sees.

Hits and misses are request counts taken from CDN logs or analytics for the same period. Total traffic is the data delivered to users in that period.

The principle

A cache works because popularity is uneven: a small set of objects (logos, scripts, popular pages and videos) accounts for most requests. Once an object is stored at the edge and still fresh under its HTTP caching rules (Cache-Control max-age, Expires, validators), repeat requests are answered locally. Only first requests, expired objects and uncacheable responses go back to the origin. The origin traffic estimate assumes that misses are, on average, the same size as hits, so the share of bytes equals the share of requests. That is a simplification: the byte hit ratio can be quite different from the request hit ratio when large files behave differently from small ones.

Worked example

Inputs: 950,000 hits, 50,000 misses, 10,000 GB total traffic.
Total requests: 950,000 + 50,000 = 1,000,000.
Hit ratio: 950,000 ÷ 1,000,000 × 100 = 95%.
Miss share: 50,000 ÷ 1,000,000 = 0.05.
Origin traffic: 10,000 × 0.05 = 500 GB.

Each point is the calculator re-run with only “Cache hits” changed; the gold point is the default example. Change the inputs above and this chart follows your numbers.

Cache hits	Hit ratio
237,500	82.61%
475,000	90.48%
712,500	93.44%
950,000	95%
1.19e+06	95.96%
1.42e+06	96.61%
1.9e+06	97.44%

How to read the result

Higher is better. Static assets with long cache lifetimes often reach very high hit ratios, while sites with personalised HTML, many query-string variants or short TTLs score much lower. Small changes near the top matter a lot: going from 95% to 99% cuts origin requests from 5% to 1%, a fivefold reduction in origin load.

Common mistakes

Comparing request hit ratio from one tool with byte hit ratio from another.
Counting requests for uncacheable content (APIs, logged-in pages) and expecting a high ratio.
Busting the cache with random query strings or cookies in the cache key.

Keep in mind: The origin traffic figure assumes equal average object size for hits and misses. Use the CDN's own byte-based origin metrics for billing or capacity decisions.

Frequently asked questions

What is a good cache hit ratio?

It depends on the content mix. For mostly static files, above 90% is common; for dynamic sites, improving the ratio for static assets is usually the realistic goal.

How can I raise the hit ratio?

Set longer Cache-Control lifetimes for versioned assets, normalise or ignore unnecessary query strings and cookies, and consider an origin shield tier.

Do revalidations count as hits?

It depends on the provider. A conditional request answered with 304 Not Modified still contacts the origin, though it transfers little data.

Sources: IETF RFC 9111, HTTP Caching

Checked 8 Nov 2026. Standard formulas and RFC/IEEE definitions, computed locally in your browser — nothing you type is sent anywhere.

Your numbers