Benchmark report
Is vercel.com slow?
Homepage load times from 5 cities
Fast globally
Vercel responds quickly from all 5 test locations. San Francisco was fastest at 64 ms; Amsterdam took 98 ms.
- Fastest
- 64 ms
- San Francisco
- Slowest
- 98 ms
- Amsterdam
- Successful requests
- 100%
- 5 of 6 locations
Response time around the world
San Francisco64 ms
Tokyo64 ms
Montreal66 ms
Singapore69 ms
Amsterdam98 ms
MumbaiCollecting since 7 Sept 2026
The slow days are still in the numbers
In the controlled run of 7 September the median request in every city finished in 47 to 81 ms: connecting took 2 to 4 ms, the TLS handshake 6 to 16 ms, the wait for the first byte 33 to 50 ms and the download 3 to 12 ms. The daily runs tell two stories. From 30 August to 3 September every city was slower, Singapore most of all with daily medians of 285 to 406 ms and Tokyo at 284 to 484 ms; since 4 September all five cities have answered in 40 to 98 ms. Those five slow days are still part of the pooled typical times. The slower times sit far above them because they are the median of two controlled runs, and one of those runs was on 30 August, inside the slow period.
Click the location to see each stage.
Amsterdam81 ms
Amsterdam- Finding the server
- 3 ms
- Reaching the server
- 3 ms
- Setting up security
- 16 ms
- Waiting for the server: most of the time
- 50 ms
- Receiving the response
- 12 ms
- Total
- 81 ms
Finding the server is measured once per location, so it sits outside these bars. Open a city to see it.
Bars show the typical request, so the parts add up to its total. Why
Compare all locationsHide the comparison
Click a city to see its stage-by-stage breakdown.
Amsterdam81 ms
Amsterdam- Finding the server
- 3 ms
- Reaching the server
- 3 ms
- Setting up security
- 16 ms
- Waiting for the server: most of the time
- 50 ms
- Receiving the response
- 12 ms
- Total
- 81 ms
San Francisco53 ms
San Francisco- Finding the server
- 10 ms
- Reaching the server
- 4 ms
- Setting up security
- 9 ms
- Waiting for the server: most of the time
- 33 ms
- Receiving the response
- 7 ms
- Total
- 53 ms
Montreal47 ms
Montreal- Finding the server
- 10 ms
- Reaching the server
- 2 ms
- Setting up security
- 7 ms
- Waiting for the server: most of the time
- 35 ms
- Receiving the response
- 3 ms
- Total
- 47 ms
Singapore50 ms
Singapore- Finding the server
- 3 ms
- Reaching the server
- 2 ms
- Setting up security
- 7 ms
- Waiting for the server: most of the time
- 34 ms
- Receiving the response
- 7 ms
- Total
- 50 ms
Tokyo50 ms
Tokyo- Finding the server
- 1 ms
- Reaching the server
- 3 ms
- Setting up security
- 6 ms
- Waiting for the server: most of the time
- 36 ms
- Receiving the response
- 5 ms
- Total
- 50 ms
The phases come from the controlled run of 7 September 2026. Typical times pool daily requests; slower times use controlled full runs. 5 of 6 test locations included: Amsterdam, San Francisco, Montreal, Singapore, Tokyo.
Finding the server is measured once per location, so it sits outside these bars. Open a city to see it.
Bars show the typical request, so the parts add up to its total. Why
Vercel still feels slow?
- Check Vercel’s status page
- Platform and edge incidents are posted there before they show up in a benchmark.
- This is not a benchmark of your deployment
- A site on Vercel has its own build output, caching rules and functions. Functions, the dashboard and the build pipeline are not measured here.
- Served from Vercel’s edge
- The homepage came from four different addresses across our cities in the controlled run, 34 ms apart at most, the pattern of an edge network. Which edge serves your visitors depends on where they are.
Response time over the last 30 days
A dot marks a day well above that location’s usual level.
Show the numbers
| Day | Amsterdam | San Francisco | Montreal | Singapore | Tokyo | Mumbai |
|---|---|---|---|---|---|---|
| 27 August 2026 | 107 ms | 69 ms | 48 ms | 58 ms | 62 ms | — |
| 28 August 2026 | 246 ms ▲ | 202 ms ▲ | 168 ms ▲ | 546 ms ▲ | 274 ms ▲ | — |
| 29 August 2026 | 234 ms ▲ | 201 ms ▲ | 138 ms ▲ | 567 ms ▲ | 267 ms ▲ | — |
| 30 August 2026 | 279 ms ▲ | 236 ms ▲ | 167 ms ▲ | 406 ms ▲ | 484 ms ▲ | — |
| 31 August 2026 | 238 ms ▲ | 189 ms ▲ | 153 ms ▲ | 385 ms ▲ | 284 ms ▲ | — |
| 1 September 2026 | 242 ms ▲ | 177 ms ▲ | 148 ms ▲ | 353 ms ▲ | 304 ms ▲ | — |
| 2 September 2026 | 185 ms | 124 ms | 76 ms | 285 ms ▲ | 344 ms ▲ | — |
| 3 September 2026 | 180 ms | 130 ms | 83 ms | 299 ms ▲ | 480 ms ▲ | — |
| 4 September 2026 | 98 ms | 56 ms | 69 ms | 58 ms | 53 ms | — |
| 5 September 2026 | 79 ms | 60 ms | 49 ms | 60 ms | 49 ms | — |
| 6 September 2026 | 76 ms | 65 ms | 66 ms | 56 ms | 52 ms | — |
| 7 September 2026 | 84 ms | 58 ms | 60 ms | 56 ms | 40 ms | 51 ms |
| 8 September 2026 | 69 ms | 55 ms | 48 ms | 90 ms | 55 ms | 44 ms |
| 9 September 2026 | 77 ms | 53 ms | 62 ms | 55 ms | 57 ms | 42 ms |
| 10 September 2026 | 85 ms | 55 ms | 51 ms | 48 ms | 52 ms | 61 ms |
| 11 September 2026 | 82 ms | 53 ms | 54 ms | 45 ms | 47 ms | 48 ms |
| 12 September 2026 | 75 ms | 64 ms | 46 ms | 50 ms | 46 ms | 46 ms |
About this measurement
GET vercel.com · 70 requests per location over 14 daily runs · 30 August 2026 to 12 September 2026
What we tested · Marketing website. The vercel.com marketing homepage, about 520 KB of HTML, served from Vercel's own cache.
How LatencyRadar measures response time →
Technical details
Doesn’t measure: A site you deploy on Vercel, its serverless functions, or the dashboard.
One Vercel surface: the vercel.com marketing homepage, about 520 KB of HTML that Vercel serves from its own edge cache. It is the page Vercel publishes on its own network, so it shows how Vercel's edge answers a cold request from each of our cities. It does not stand for your project: a site you deploy on Vercel has its own build output, caching rules and functions, and its speed depends on those choices as much as on the network.
Not measured: sites deployed on Vercel, serverless or edge functions, the dashboard, image optimisation, or anything a browser does after the HTML arrives. We cannot see Vercel's cache state from outside, so a slow day may be a cache miss, a deploy, or the path between our worker and the nearest edge; the numbers do not say which.
Between 30 August and 12 September 2026 the typical time ranged from 64 ms in San Francisco and Tokyo to 98 ms in Amsterdam. Those numbers hide a split. From 30 August to 3 September every city was slower, and Asia most: Singapore's daily median ran from 285 to 406 ms and Tokyo's from 284 to 484 ms. Since 4 September all five cities have answered in 40 to 98 ms, and the controlled run of 30 August (p95 of 508 to 584 ms in Amsterdam, Singapore and Tokyo) against 7 September (111 to 187 ms) shows the same shift. We cannot see from outside whether that was a cache, a deploy or a routing change. The slower times on this page are the median of those two runs, so they still carry the slow one. Netlify and Fastly, measured the same way on the same days, did not move.
| Test location | Typical response time | Slower response time | Requests |
|---|---|---|---|
| Amsterdam | 98 ms | 313 ms | 20 |
| San Francisco | 64 ms | 214 ms | 20 |
| Montreal | 66 ms | 127 ms | 20 |
| Singapore | 69 ms | 348 ms | 20 |
| Tokyo | 64 ms | 368 ms | 20 |
| Mumbai | Collecting since 7 Sept 2026 | ||
Typical: half of the requests finished within this time (technical: p50). Slower: 95% of requests finished within this time (technical: p95). Statistics
- Request
- GET vercel.com
- Measured from
- Amsterdam · San Francisco · Montreal · Singapore · Tokyo
- Requests
- Daily measurements from 30 August 2026 to 12 September 2026. 5 of 6 test locations included: Amsterdam, San Francisco, Montreal, Singapore, Tokyo. · 30 August 2026 to 12 September 2026
- Timings taken
- DNS, connect, TLS, waiting for the server, download
- Report coverage
- 5 of 6 test locations included: Amsterdam, San Francisco, Montreal, Singapore, Tokyo.
- Daily requests in each location
- Amsterdam: 70 · San Francisco: 70 · Montreal: 70 · Singapore: 70 · Tokyo: 70
- Typical response time
- Median of the eligible daily requests in each included location.
- Slower response time
- Median of recent controlled full-run p95 values in each included location.
Vercel
https://vercel.com/
Independent measurement by LatencyRadar. Not affiliated with Vercel. · How we measure (v1.1)
How fast does your site load around the world?
Run a free speed test from multiple cities and find out where your users are waiting. No setup, no account required.
No account required · Takes about 30 seconds.
