How much of your Sentry bill is noise?

Connect Sentry with a read-only token. We read your last 90 days of issues, estimate what the noise costs you, and hand you the exact ignoreErrors snippet and inbound filters to stop it. Free.

Read-only scopes: org:read, project:read, event:read. We never read your source code, and we never store event payloads.

Example reportSample data
Estimated noise in the last 90 days
38% of your events
roughly $120 a month of quota, estimated
ignoreErrors snippet
Copy into your SDK config
Inbound filter checklist
Toggle in Sentry settings
Archive list
Every issue, with a reason

Your report is built from your own Sentry issues. The dollar figure is an estimate from an editable per-event rate, not a copy of your invoice.

The problem

You are paying per event, including the useless ones

The audit does the arithmetic nobody does by hand: noise events times what you pay per event, projected over a month.

Noise bills like signal

Sentry charges per error event. An extension crash loop and a real payment failure cost exactly the same. Every event you would never act on is quota you already paid for.

It is the same few sources

Browser extensions, headless browsers and crawlers, local development addresses, and errors the browser itself emits: ResizeObserver loops, bare Script error, dead-object access. None of them are bugs in your code.

Nobody can see how much

Sentry tells you how many events you sent, not how many were worth sending. Without that split there is no way to know whether to raise your quota or cut the noise.

How it works

Connect, classify, fix

Three steps, and the first report costs nothing.

01

Connect, read-only

Paste a Sentry auth token with org:read, project:read and event:read, then pick the projects to audit. The token is encrypted before it is stored. Archiving later asks for a separate write token, so the audit itself never gets write access.

02

We classify, with reasons

We import your top issues by event count over the last 90 days, then sort them: deterministic rules first, volume and staleness signals next, and an AI pass only on what is left. Every verdict carries the reason it was made, and you can override any of them.

03

You fix it

Get a generated ignoreErrors block, an inbound-filter checklist, and the archive list. Copy them for free. On Pro, archive the noise in Sentry in one click, with undo.

What you get

A report you can act on the same afternoon

The audit ends in copy-paste artifacts, not a dashboard to check. All of them are free.

Generated ignoreErrors snippet
// Generated by NoiseGate from your last audit.
Sentry.init({
  dsn: "YOUR_SENTRY_DSN",
  ignoreErrors: [
    "ResizeObserver loop limit exceeded",
    "Script error.",
    /chrome-extension:\/\//i,
  ],
});

Example output. Your entries come from the issues your own audit classified as noise.

Inbound filter checklist

The built-in Sentry filters worth turning on, as a checklist for Settings, Inbound Filters. No code change, and it drops the events before they are billed.

Archive list

Every issue we called noise, with its Sentry id, its event count and the reason for the verdict. Free to copy. On Pro it becomes a one-click bulk archive.

Reversible, always

Nothing is changed in your Sentry unless you click archive, and archiving has an undo. Remove a filter or an ignoreErrors entry and those events flow again.

Pricing

Start free, scale when ready

The audit is free, for good. Pro is $29 a month, or $290 a year, and pays for archiving and monitoring.

Free

$0

The whole audit, and everything you need to act on it

  • Full Sentry noise audit and report
  • The ignoreErrors snippet, inbound filters and archive list
  • Re-scan once a day
Run your audit
Popular

Pro

$29

For teams that want the noise to stay gone

  • Everything in Free
  • One-click archive in Sentry, with undo
  • Weekly monitoring and drift emails
  • Unlimited re-scans
Start 7-day trial

FAQ

Frequently asked questions

Do you have any questions? We have got you covered.