Redirect speed benchmark
Updated hourly · last 7 days · 1 URL measured

How fast is Buffer?

Buffer redirects answered in a median 40 ms from Washington, D.C. over the last 7 days. That is faster than 33% of all redirects tested here.

Median first byte
40 ms
p90 43 ms
Rank
#14
of 25 providers
Faster than
33%
of tested redirects
vs RedirHub
+25 ms
2.7× RedirHub's time

By location

Where Buffer is fast, and where it isn't

Distribution

Buffer among every redirect tested

RedirHub 15
all 85
Buffer 40
0200400600750+ ms

Bars: median first-byte time of every URL tested from Washington, D.C. in the last 7 days. Lines: provider medians.

All providers

Redirect speed by provider, last 7 days

Median first byte from Washington, D.C. over the last 7 days. Every provider with at least one tested URL is listed, with how many URLs its figures come from.

#ProviderMedianURLs
1Fastly11 ms5
2Amazon CloudFront11 ms1
3RedirHubUs15 ms2
4Short.io15 ms2
5Microsoft Azure15 ms1
6Google17 ms1
7Meta19 ms1
8Bitly21 ms2
9Amazon25 ms1
10Envoy29 ms3
11Cloudflare31 ms3
12Rebrandly32 ms1
13Dub37 ms1
14Buffer40 ms1
15Akamai41 ms3
16LinkedIn62 ms1
17TinyURL68 ms2
18Vercel75 ms1
19YouTube79 ms1
20redirect.pizza166 ms2
21Spotify184 ms1
22urllo196 ms2
23BL.INK239 ms2
24Cuttly333 ms1
25nginx341 ms4

Buffer links we measure every day

Open one for its full report: every hop, its status code and the time of each phase, from all three locations.

Where the time goes
Buffer
40 ms
RedirHub
15 ms
DNSConnectTLSServer wait
Moving from Buffer?

Bring your links to your own domain. RedirHub serves redirects from 130 Global PoPs, with a ~90ms global average response time.

Start with your domain →

Method

How these numbers are measured

A fixed list, every day

Well-known redirects and short links are measured every day from Washington, D.C., Frankfurt and Tokyo, so every location always has a baseline.

Plus what people test

Every URL tested on the speed test counts once per location, grouped by the provider that answered it. Entries expire after 7 days, and refused answers (403, 429) are left out.

Same measurement for all

Fresh connection, no redirect following, median of every run. Run the same test with curl →