Skip to content
Comparison

Umami vs Plausible: which one to pick

Umami or Plausible? Licence, hosting, cookies, features and integration compared, with server-side measurement as a third option for PHP sites.

Recorded on November 4, 2026

We publish Quiet Metrics, which is neither Umami nor Plausible. We add it as a third column because it addresses a limit the two share, and we also say when it does not fit. Everything written here about Umami and Plausible comes from their own published pages, dated and linked at the bottom.

At a glance

Umami Plausible Quiet Metrics (server-side)
Licence MIT Community Edition under AGPLv3, script under MIT Closed platform; PHP packages under MIT
Self-hosting Node.js and PostgreSQL Elixir, PostgreSQL and ClickHouse Not offered
Cloud hosting Servers in the US and in the EU EU, server in Germany France, at OVH
Cookies None, according to Umami No cookies and no local storage, according to Plausible No identifying or tracking cookies; a ten-minute visit continuity cookie with no identifier
How visitors are counted Method not described on the pages we read Daily fingerprint, salt deleted every 24 hours Daily fingerprint, salt destroyed every night
Script weight Not announced 54 times smaller than GA4's, according to Plausible Under 4 KB compressed, none for server-side measurement
Sending from a server HTTP API, call to write yourself Events API, call to write yourself 3 Composer packages (PHP, Laravel, Symfony)
Beyond audience Funnels, journeys, retention, heatmaps, session replay, revenue In the cloud: funnels, journeys, revenue goals Goals and campaigns, no behavioural analysis
MCP server Official, read-only Community-built, by the Sentry team Our own, read-only, from the Pro plan
SEO issues ranked by your traffic Not announced Not announced; Google Search Console integration Yes: built-in crawler, ranked by each page's views and conversions

Umami in brief

Umami is an open source analytics tool under the MIT licence, which describes itself as a web and product analytics platform. It installs with Node.js and PostgreSQL, or runs in the cloud. It says it uses no cookies, and its scope is broad: funnels, journeys, retention, heatmaps, session replay, revenue linked to your payments.

Plausible in brief

Plausible is a European analytics tool whose Community Edition is published under AGPLv3. It says it sets no cookies and uses no local storage, and it publishes its counting method: a daily fingerprint built from a salt deleted every 24 hours, the domain, the IP address and the browser, without storing the raw address or browser string. Its data is processed in the European Union, on infrastructure owned by European companies.

The real differences between the two

The licence. MIT allows almost anything, including embedding the code in a closed product, as long as the licence notice travels with it. AGPLv3 requires you to publish your changes if you run a modified version for other users over a network. For internal use without changes, the difference is small.

The stack to maintain. Umami fits in one application and one PostgreSQL database. Plausible's Community Edition adds ClickHouse to PostgreSQL: one more database to monitor, back up and update.

Where cloud data lives. Umami Cloud has servers in the United States and in the European Union. Plausible processes everything in the European Union, with a server in Germany. If your requirements exclude the United States, check the region before choosing Umami Cloud.

How open the counting is. Plausible publishes its formula. The Umami pages we read do not describe theirs.

Feature depth. Umami goes further, with heatmaps, session replay and retention. Plausible stays focused on audience, and its Community Edition states its limits: no funnels, journeys, revenue goals, single sign-on or sites API, basic bot filtering, and two releases a year. At Umami, email reports and the streaming API are cloud-only.

The MCP server. Umami maintains an official read-only server. Plausible's documentation points to a community server built by the Sentry team, which needs a key for their Stats API.

What they share: the browser

Both measure through a script loaded in the browser, and what the browser loads, the browser can block. Plausible itself estimates that 5 to 25 % of visitors with strict blockers go uncounted without a proxy, depending on the audience. Both document a proxy through your own domain to work around it, with its own share of setup.

Both also accept events sent from a server, provided you pass on the visitor's User-Agent. Neither documents a PHP library: the call is yours to write and maintain.

The third option: measuring from the server

If your site runs on PHP, measurement can leave from your application rather than from the browser. That is what Quiet Metrics does, with three Composer packages:

composer require quiet-metrics/laravel-metrics

// bootstrap/app.php: every pageview in the web group leaves from the server
->withMiddleware(function (Middleware $middleware) {
    $middleware->appendToGroup('web', \QuietMetrics\Laravel\Middleware\TrackPageview::class);
})

There is nothing to block on the client, because there is nothing on the client. Two limits, to be complete. A page served by a cache in front of the application does not run PHP and is not counted: add our script there, and the two methods coexist without double counting. And we offer neither self-hosting nor open source for the platform. Our counting method is published on how we count, and server-side tracking explains the approach.

Which one to pick

Pick Umami if you want to self-host with the shorter stack of the two, an MIT licence, or observation features such as heatmaps and session replay.

Pick Plausible if you want a published counting method, data processed in the European Union only, and a tool focused on audience.

Look at server-side measurement if your site runs on PHP, Laravel or Symfony, ad blockers cost you visits, and you do not want to run a server.

Do not pick us if self-hosting is a requirement, your stack is not PHP, or you need behavioural analysis. The detailed comparisons are on Umami alternative and Plausible alternative, and all of them on our comparisons page.

Frequently asked questions

Umami or Plausible: which is more privacy-friendly?

Both say they use no cookies and do not track visitors across sites. Plausible also publishes its counting method and processes data in the European Union only. With Umami, the location depends on the cloud region or on your own server.

Can Umami and Plausible be self-hosted?

Yes, both. Umami needs Node.js and PostgreSQL. Plausible's Community Edition needs PostgreSQL and ClickHouse, and lacks some cloud features.

Do Umami and Plausible have funnels?

Umami does. Plausible does in the cloud, but not in its Community Edition. Quiet Metrics does not.

Do I need a cookie banner with Umami or Plausible?

Plausible says its tool does not need one. Umami says the answer depends on the other services your site uses. In both cases, as with us, your purposes and your setup decide.

Which one is easier to integrate into a PHP application?

For both, the usual path is a script in the <head>. To measure from PHP, you write the HTTP call yourself. Our Composer packages do that sending for you: after the response with the Laravel middleware or the Symfony bundle, during the request, within a 400 ms budget by default, with the plain PHP SDK. If our service does not answer, the page is still served.

Sources and verification date

Recorded on 4 November 2026 from pages published by Umami:

Recorded the same day from pages published by Plausible:

Checked on 4 November 2026, for SEO health:

We copy no prices: they depend on volume and change over time. Their own pages are authoritative.

Umami and Plausible are trademarks of their publishers. This comparison is made independently, from public information.

Measure a first site.
No identifying or tracking cookie.

Create an account See pricing