Umami

Open-source, self-hostable, cookieless. The free Plausible if you can run a Postgres.

VISIT UMAMI

Key takeaway: If you already run a Postgres and a deploy target, Umami gets you Plausible-shaped analytics for the cost of infrastructure you are paying for anyway. It is the pick for developers who would rather own a small deployment than a subscription. Non-developers, and anyone who wants someone else responsible for uptime, should pay for Plausible.

Quick facts

  • CategoryOpen-source / self-host
  • PricingOpen source
  • Starts atFree (self-host)
  • LicenseMIT
  • Created2020
  • GitHub stars25k
  • Script size~2KB
  • CookielessYes
  • Self-hostYes

What it is

Umami is the open-source self-hosted answer to Plausible. MIT-licensed, deploys to Vercel + Postgres in 5 minutes, dashboard is clean, cookieless by default. A managed cloud tier exists (Umami Cloud) for teams who do not want to self-host. Active development, healthy GitHub community.

Best for

  • Developers comfortable running Postgres + Vercel who want zero ongoing cost
  • Teams running multiple sites who want one self-hosted dashboard
  • Anyone who wants Plausible's philosophy on a permissive license

When not to pick it

Skip Umami if you do not want to maintain self-hosted infrastructure. The managed Cloud tier exists but is less mature than Plausible Cloud.

My take

Umami is the right default for developers who already run a Postgres and want to drop $9-15/mo for analytics. Deploys to Vercel free tier + Supabase free tier and just works. For non-developers, Plausible is the better managed answer.

The deploy is fast, the ownership is permanent

Umami is a Next.js app plus a database schema, which is the whole reason it took over this niche. Point it at Postgres, set a couple of environment variables, deploy, paste a script tag. There is no separate columnar store to keep alive next to it, which is the real difference from self-hosting Plausible, where ClickHouse comes with the box. MIT licensing means no copyleft obligation, so agencies can run it inside a client stack without a legal conversation first. What you own afterwards is a schema that grows with pageviews, upgrades you apply yourself, and a database you back up. On a low-traffic marketing site that is close to zero work. At millions of events a month, Postgres becomes the thing you are tuning, and that is the point where a ClickHouse-backed tool starts to look sensible again.

One dashboard, many sites, and the Cloud alternative

The strongest practical case for Umami is a portfolio. An agency or solo operator running twenty properties finds per-site subscription pricing turns silly quickly. One Umami deployment handles all of them, each with its own website ID, and the marginal cost of the twenty-first site is a row in a table. Umami Cloud exists for teams that want the product without the deployment, though it is less mature than Plausible Cloud and has less track record behind it. The dashboard itself is clean and covers pageviews, referrers, devices, countries, and custom events, but it stops well short of funnels and retention. Treat it as the free-to-run counterpart to Plausible rather than a step toward product analytics. PostHog is the tool for that half of the job.

Frequently asked questions

Is Umami really free?

The software is, under the MIT licence. You pay for whatever runs it: a host for the Next.js app and a database, both of which fit comfortably on small or free tiers for a low-traffic site. Your real cost is maintenance, upgrades, and backups. A managed Umami Cloud tier exists if you would rather not own that.

Umami or Plausible?

Umami if you are a developer who already operates a database and wants no recurring bill. Plausible if you want somebody else accountable for uptime. Feature-wise they are close: both cookieless, both light, both web-analytics-only. Plausible's cloud product is more mature, and its self-hosted build needs ClickHouse, which is heavier to run than Umami's plain SQL database.

Can Umami handle high traffic?

Up to a point. Postgres is fine for typical marketing-site and small-SaaS volumes, but it is a row store handling an append-heavy analytics workload, so ingest and query cost climb with event count. At high volume expect to tune indexes, prune old data, or move to a ClickHouse-backed tool. Plan that exit before you need it.

Links

Compare Umami side-by-side

Similar tools you should also consider

If Umami is your pick, the next conversation is short

The 30-min call is where your analytics choice becomes a real implementation, an event-tracking plan that fits your product, and a privacy story that does not need a 4-screen cookie banner.