Slowpoke compared with the alternatives

Every page here says plainly what the other tool does better, because a comparison that only flatters us is worth nothing to someone deciding what to install this week. Slowpoke is free and self-hosted, so there is no sale to protect.

What it costs

The table nobody selling you monitoring will print

Public list prices, read in September 2026, for three shapes of team. Every figure is an estimate and real invoices carry discounts and commitments — and where a vendor's tiers are not public enough to compute a number honestly, the cell says what the meter is instead of inventing one. That restraint is the whole point of the table.

Tool, and what it meters	Small3 servers · 1M requests a month · 5 engineers	Growing12 servers · 10M requests a month · 15 engineers	Bigger40 servers · 50M requests a month · 40 engineers	The detail
SlowpokeNothing is metered	€0	€0	€0	Open source and self-hosted. You pay for the machines you already run.
Datadog APMPer host, counted hourly and billed at the peak	≈ $1,100 a year	≈ $4,500 a year	≈ $14,900 a year	At about $31 per host per month on an annual commitment. APM only: infrastructure, logs, RUM and synthetics are separate products, and ingested spans beyond the included allowance are extra.
New RelicPer gigabyte ingested, plus a licence per person	≈ $2,400 a year	≈ $28,700 a year	≈ $75,800 a year	At list: Full Platform users about $349 a month each, Core users about $49, data at $0.40 a gigabyte, with 100 GB and one full user free. Assumes 1/5/12 full users and 4/10/28 core users, and 50 GB, 500 GB and 2 TB of ingest.
Sentry, self-hostedNo licence; you pay in infrastructure	€0 + the fleet	€0 + the fleet	€0 + the fleet	Self-hosted Sentry is free to run and is a fleet: Relay, Kafka, ClickHouse and Postgres. The cost is the servers and the person who keeps it alive.
AppSignalPer request — a web request or a background job	from ≈ $279 a year	a higher request tier	a higher request tier	From about $23 a month, roughly $279 a year at 250,000 requests a month. Unlimited apps, hosts and users, which is unusually fair. Larger tiers are not public enough to compute honestly.
Scout MonitoringTransaction tiers, plus log storage	from ≈ $19 a month	named plans from ≈ $161 a month	named plans from ≈ $161 a month	Build-your-own tiers start around $19 a month; Basic, Plus and Pro list at roughly $161, $249 and $499. The driver people underestimate is logs: 1 GB included, about $40 a month per additional 30 GB.
Laravel NightwatchPer event — every query, job and cache hit	€0 under 300k events	≈ $40–80 a month at 1M+ events	scales with events	One ordinary request can emit a dozen events, so event volume runs well ahead of request volume. Overage is about $0.35 per 100,000 on the lower paid tiers.
TidewaysPer PHP request, monthly limit	from ≈ €29 a month	a higher request tier	a higher request tier	Team from about €29 a month at 250,000 requests, Business from about €89, and a Flex plan from about €69 with unlimited applications under one shared request limit.
BlackfirePer trace quota, from one million a month	a traces quota	a traces quota	a traces quota	A profiler rather than a monitor, so traces are taken deliberately rather than continuously. Public pricing starts at one million traces a month.
SigNoz, self-hostedNo licence; you pay for ClickHouse	€0 + ClickHouse	€0 + ClickHouse	€0 + ClickHouse	The open source core is free. At volume the storage backend, not the tool, decides the bill — and somebody has to run it.
Coroot, self-hostedNo licence for the open source core	€0	€0	€0	Genuinely flat at volume, which is one of its better arguments. Needs privileged eBPF and is happiest on Kubernetes.
Percona PMMNo licence	€0	€0	€0	Free and self-hosted, and the server bundles Grafana, a metrics store and a query analytics backend. Databases only.
Elastic APMNo licence for the free tier; you pay for the cluster	€0 + Elasticsearch	€0 + Elasticsearch	€0 + Elasticsearch	Close to free if the Elastic Stack is already in the building. A search cluster to operate if it is not.
Grafana stackNo licence; you pay in assembly and storage	€0 + the work	€0 + the work	€0 + the work	Prometheus, Loki, Tempo and Grafana cost nothing to license and a great deal to put together and keep.
Laravel PulseNothing is metered	€0	€0	€0	Free and first-party, and it writes to your application's own database. One app, live window only.
SonarQube CommunityNo licence for Community; paid editions by lines of code	€0 Community	by lines of code	by lines of code	A different category: it reads the repository in CI rather than watching production.

Slowpoke is open source and self-hosted: there is no licence, no host count, no seat count and no event count, so the column stays at zero however far right you read. What you pay for is the machines you were already paying for.

All of them at once

Sixteen tools, six questions, and what each one is genuinely best at

The last column is the honest one, and for most rows it does not say Slowpoke. A tick means the tool does this properly; a tilde means partly, or with a condition worth reading on its page.

Tool	Self-hosted	Free	Finds your estate for you	Ranks by business cost	Query plan and the fix	File and line of the query	Best at
Slowpoke	yes	yes	yes	yes	yes	yes	Deciding what to fix first, and proving the fix held
Datadog	no	no	partly	no	partly	no	Knowing what is happening right now, across everything
New Relic	no	partly	partly	no	partly	no	One broad platform, if you can pay per person
Sentry	partly	partly	no	no	no	partly	Errors: what threw, where, and for whom
AppSignal	no	no	no	no	no	partly	A hosted APM that does not charge per host or per seat
Scout Monitoring	no	no	no	no	no	yes	N+1 detection across four ORMs, with the backtrace
SigNoz	yes	yes	no	no	no	no	An OpenTelemetry-native store for traces, logs and metrics
Coroot	yes	yes	yes	no	no	no	A service map on Kubernetes with no instrumentation at all
Percona PMM	yes	yes	partly	no	yes	no	The deepest free view of a database engine
Elastic APM	yes	partly	no	no	no	no	Traces next to the logs you already keep in Elasticsearch
Grafana stack	yes	yes	no	no	no	no	Building exactly the dashboard and alert you had in mind
Blackfire	no	no	no	no	no	yes	The call graph inside one request, function by function
Tideways	no	no	no	no	no	yes	PHP monitoring and a callgraph profiler from one vendor
Laravel Nightwatch	no	partly	no	no	no	yes	The widest first-party instrumentation inside Laravel
Laravel Pulse	yes	yes	no	no	no	partly	A free live wall for one Laravel app, in five minutes
SonarQube	yes	partly	no	no	no	no	Holding the code itself to a standard, in CI

One at a time

The full comparison, tool by tool

Hosted APM platforms

They watch everything and bill you for it, on a meter of one kind or another. Slowpoke overlaps with the part you use to hunt slow endpoints and slow queries, and nothing else.

Slowpoke vs Datadog The complete platform, billed per host, against one free ordered list of what to fix. Slowpoke vs New Relic Per-gigabyte ingest plus per-seat licensing, against no seats and no ingest meter at all. Slowpoke vs Sentry Errors first, and a self-hosted install that is a fleet. Slowpoke is one binary about slowness. Slowpoke vs AppSignal A likeable, fairly priced hosted APM billed per request. Slowpoke has no meter and no cloud. Slowpoke vs Scout Monitoring N+1 detection across four ORMs is their signature — and the log bill is the real cost.

Self-hosted telemetry

The closest neighbours, and the comparisons worth reading carefully: they store and query telemetry, which is a different job from producing an ordered list of work.

Slowpoke vs SigNoz The OpenTelemetry-native store everyone recommends, against a tool that arrives with an answer. Slowpoke vs Coroot eBPF service maps with no instrumentation — the closest thing to us, and Kubernetes-shaped. Slowpoke vs Percona PMM The deepest free database monitoring there is — and it stops at the database. Slowpoke vs Elastic APM The other free, self-hostable APM — if you are willing to run an Elasticsearch cluster. Slowpoke vs the Grafana stack Prometheus, Loki, Tempo and Grafana: infinite flexibility, and weeks of assembly.

PHP profilers

Not really rivals. A profiler explains what happens inside one request; Slowpoke tells you which request is worth profiling. For a PHP team the honest answer is usually both.

Slowpoke vs Blackfire A deterministic profiler for one request. Slowpoke tells you which request to profile. Slowpoke vs Tideways PHP monitoring and profiling in one, priced per request. Slowpoke is free and not PHP-only.

Framework tools

Excellent inside one framework. Slowpoke covers the framework, the machine it runs on, the database behind it and the four other services nobody instrumented.

Slowpoke vs Laravel Nightwatch First-party and superb inside Laravel, billed per event. One request can emit a dozen. Slowpoke vs Laravel Pulse Free, first-party, and a live wall for one app. Slowpoke keeps history and makes work happen.

Code analysis and technical debt

They read your repository. Slowpoke reads the system that is actually running, and prices the debt in the time your users lose.

Slowpoke vs SonarQube Static analysis and a debt figure in hours. Slowpoke measures the debt your users actually pay.

Or just try it

One server, Docker, fifteen minutes. It takes less time than reading a comparison table.

Run it on your own server Try the live demo