Sentry is open source.
sentry.io charges to host it.
The Sentry server you're paying $5K/mo for is the same code base at github.com/getsentry/sentry — yours to run for free under the BSL. The catch: production-grade self-hosting (multi-tenant orgs, alerting integrations, retention policies, ClickHouse tuning, upgrade hygiene) is real engineering work. We deploy, configure, and operate it. Same Sentry UX, your AWS, no per-event pricing.
sentry.io Business · 50M events/mo
Math compounds with event volume: at 500M events/mo (typical 250-eng SaaS), Y1 reclaim crosses $200K.
sentry.io charges per event. ClickHouse charges per byte.
sentry.io's pricing scales with errors + transactions + replays + spans. Self-hosted Sentry on AWS scales with your storage and compute — both of which are dramatically cheaper at every volume past Hobby tier.
sentry.io list prices May 2026. Self-host numbers based on Sentry self-hosted on AWS EC2 (multi-AZ) + ClickHouse for events + Postgres for metadata + Allied $14K deploy amortised. ClickHouse storage at $0.023/GB plus compute scales much more gently than per-event SaaS pricing.
Three layers. Same Sentry, your tenancy.
The same Sentry server stack the team at sentry.io runs — Relay + Snuba + ClickHouse + Postgres + Redis — deployed in your AWS account with multi-tenant org config, retention policies, alerting integration, and SLA-backed ops.
- Drop-in compatible with sentry-sdk
- Rate limiting + sampling per project
- PII scrubbing rules
- Multi-region ingest endpoints
- Auto-scaling on AWS ECS / EKS
- ClickHouse — events / spans / replays
- Postgres — projects, users, alerts
- Redis — queues + buffers
- S3 — replay payloads + uploaded source maps
- Configurable retention per project (7 / 30 / 90 days)
- Sentry dashboard at sentry.yourcompany.com
- SSO via Okta / Azure AD / Google
- Slack / PagerDuty / OpsGenie webhooks
- Allied SLA-backed ops layer
- Quarterly upgrade cadence
Sentry is BSL 1.1 licensed — free for self-host, restrictions only on offering Sentry-as-a-service to third parties. Internal company use is unrestricted. The license respects your right to self-host; Allied stays well within its terms.
It's the same Sentry.
Self-hosted Sentry is the same code base as sentry.io. Same UI, same SDKs, same workflow. The differences are operational (multi-tenant config, retention policies, upgrade hygiene) rather than feature-level.
Fixed price. Sized by event volume.
One number, one ship date. Sized by event volume + required HA posture.
From $14,000 — fixed
- Sentry deployment on your AWS (ECS / EKS configurable)
- ClickHouse + Postgres + Redis sized for your volume
- SSO integration — Okta / Azure AD / Google Workspace
- Alerting — Slack / PagerDuty / OpsGenie webhooks configured
- Retention policies per project (configurable)
- SDK migration helpers — drop-in DSN swap
- Source map upload + symbolication workflow
- Backup + DR — daily snapshots + restore tested
- 1-week parallel run with sentry.io before cutover
- Runbook + handover doc — your team can run it solo if you choose
24-hour quote turnaround · NDA on request
Why we don't recommend the obvious alternatives.
"Sentry alternative" results: Bugsnag, Rollbar, Datadog Errors, Honeybadger. Same hosted-vendor model, different brand. Your team already knows Sentry's UX — keep it.
"What about…"
Will my SDK / DSN config break?▾
How does upgrade hygiene work?▾
What about Replay performance?▾
What if Sentry releases a new feature only on cloud?▾
What about ClickHouse — isn't that complex?▾
Who owns everything?▾
Why this number works: AI-paired senior engineers raise the floor on quality + consistency · 250+ products shipped · direct vendor billing — you keep the wrapper margin. → The full method
SCOPE · WHAT WE REBUILD vs WHAT WE DON’T · SENTRY.IO (HOSTED)
We don’t rebuild Sentry.io (hosted). We rebuild your usage of it — the 15–25% you’re paying for and actually using, on direct billing. Read the universal methodology →
- Error capture + stack traces — client + server SDKs, with breadcrumbs and user context, into your Postgres or Loki
- Release tracking — tag errors with release version, surface regression rates in dashboard
- Source map upload + symbolication — for minified JS / mobile builds, automated in CI / CD
- Alert rules — your existing rules (volume thresholds, regression spikes) wired to Slack + PagerDuty
- Issue triage UI — group similar errors, assign owners, comment threads, resolve workflow
- Basic performance traces — transaction sampling for p50 / p95 / p99 latency on your top 10 routes
- Sentry session replay — specialized. Our pick: FullStory or LogRocket direct (more polished UX)
- Sentry profiling (CPU / memory) — specialized feature. Our pick: stay on Sentry for this slice or use language-specific tools (py-spy, pprof)
- Sentry Codecov integration — different product. Our pick: Codecov direct (you may already pay them) or Coveralls
- Sentry Cron monitoring — niche feature. Our pick: Healthchecks.io (much cheaper, better UX)
- Sentry metric alerts — niche. We wire metrics to Grafana + Alertmanager direct (covered in our Datadog rebuild)
Not a rebuild case? We still help. Free 30-min advisory call to evaluate any of the alternatives above — we use them all in our practice and have no resale relationships, just opinions earned from the work. Book a call →
Stop the per-event pricing. Self-host the OSS.
From $14,000 fixed. 3–4 weeks to ship. Same Sentry UX. Your AWS, your data, your event volume scales without per-event pricing. No deck-pitch — send your sentry.io plan + monthly event volume, we send a quote in 24 hours.
15+ years · ISO 9001:2015 · 250+ shipped · MSA · SOW · DPA ready · BSL-compliant deployment.