Sentry costs $0 on Developer, $26 a month on Team and $80 a month on Business, all billed annually, with Enterprise quoted. The part that surprises people is that Team and Business include exactly the same allowances: 50,000 errors, 5 million spans, 50 replays, 5 GB of logs, 5 GB of application metrics, one uptime monitor and one cron monitor on both. The $54 a month between them, $648 a year, buys SSO, audit logs, advanced quota management and unlimited dashboards. It buys no additional volume at all. Real Sentry bills are therefore decided almost entirely by the meters that sit underneath the plan, not by which plan you pick.
Every figure below was read from sentry.io/pricing. The arithmetic is ours and each step is shown, so you can substitute your own volumes. One honest caveat up front: Sentry's monthly-billed prices sit behind a toggle that is not in the served HTML, so the plan prices here are the annual-billed rates the static page states. Monthly billing costs more and you should read the toggle yourself before budgeting.
The four Sentry plans in 2026
Sentry sells one subscription that covers error monitoring, tracing, session replay, logs, profiling, uptime and crons. There are no separate products to buy, which is genuinely simpler than most of this market. What there is instead is a long list of meters.
| Plan | Price | Errors | Spans | Replays | Logs | Uptime | Crons | Dashboards | Retention |
|---|---|---|---|---|---|---|---|---|---|
| Developer | $0 | 5,000 | 5M | 50 | 5 GB | 1 | 1 | 10 | 30 day lookback |
| Team | $26/mo | 50,000 | 5M | 50 | 5 GB | 1 | 1 | 20 | Up to 90 day |
| Business | $80/mo | 50,000 | 5M | 50 | 5 GB | 1 | 1 | Unlimited | Up to 90 day |
| Enterprise | Quoted | Custom | Custom | Custom | Custom | Custom | Custom | Unlimited | Custom |
The Team to Business jump is 3.08x the price for zero extra volume
Read the Team and Business rows side by side and the only cell that changes is the dashboard count. Business costs $80 against Team's $26, which is 3.08 times the price, a 208% increase, for an identical allowance of every metered thing Sentry sells. That is not a criticism, it is just a fact worth knowing before you upgrade: you are buying governance, not headroom. SSO, audit logs and advanced quota management are exactly what a company with a security review needs, and $648 a year is a reasonable price for them. But if you are upgrading because you keep running out of error quota, Business will not help. Buying reserved volume on Team will.
What Sentry costs beyond the allowance
This is the part of the rate card that decides your invoice. Sentry gives you two ways to exceed an allowance: reserve volume up front at a discount, or let it run and pay as you go. Sentry does not publish a headline discount percentage for reserving, so the only honest way to size it is from its own published unit rates: on the Team error tier above the 50,000 allowance, reserved volume is $0.00029 per error against $0.00036 pay as you go. That is a 19.4% saving, or put the other way, paying as you go costs 24.1% more than reserving the same volume.
| Meter | Pay as you go rate | What that means in practice |
|---|---|---|
| Logs | $0.50 per GB | 50 GB a month is $22.50, 250 GB is $122.50 |
| Application metrics | $0.50 per GB | Same rate as logs, separate 5 GB allowance |
| UI profiling | $0.25 per hour | A full 730 hour month on one service is $182.50 |
| Continuous profiling | $0.0315 per hour | The same 730 hours is $23.00 |
| Uptime alerts | $1.00 per alert | 20 endpoints is $19.00 a month on top |
| Cron monitors | $0.78 per monitor | 10 scheduled jobs is $7.02 a month on top |
UI profiling costs 7.94 times what continuous profiling costs
Two profiling meters sit on the same rate card at $0.25 and $0.0315 an hour, a ratio of 7.94 to one. They measure different things, UI profiling covers browser and mobile sessions while continuous profiling covers backend services, so this is not an apples to apples overcharge. It does mean that leaving UI profiling on at full sample rate across a busy frontend is the single fastest way to turn a $26 subscription into a three figure invoice. Sample rates are the control here, and they are worth setting deliberately rather than accepting the default.
Every plan includes exactly one uptime monitor and one cron monitor
This is the line that catches teams who assumed the paid plan covered availability monitoring. It does not. One uptime monitor and one cron monitor are included on Developer, on Team and on Business alike, and everything past the first is metered. A modest production estate makes that visible fast.
| What you watch | Chargeable uptime | Chargeable crons | Add-on per month | Add-on per year |
|---|---|---|---|---|
| 12 endpoints, 8 jobs | 11 | 7 | $16.46 | $197.52 |
| 20 endpoints, 10 jobs | 19 | 9 | $26.02 | $312.24 |
| 35 endpoints, 25 jobs | 34 | 24 | $52.72 | $632.64 |
The middle row is the one to sit with. Nineteen extra uptime alerts cost $19.00 a month, which happens to be what an entire flat-rate monitoring plan costs elsewhere, and none of it replaces the Sentry subscription underneath. There is a fuller breakdown of the interval, the three-failure threshold and what the add-on actually buys in our page on Sentry uptime monitoring and cron monitoring.
Three worked Sentry bills
Rate cards are abstract. Here are three realistic accounts with the arithmetic shown, so you can find the one closest to your own shape.
A solo developer on the free Developer plan
Within the 5,000 error and 5 GB log allowances, but watching three URLs and two scheduled jobs. Two chargeable uptime alerts at $1.00 and one chargeable cron monitor at $0.78 comes to $2.78 a month, $33.36 a year. For a side project that is close to unbeatable, and it is the case where Sentry alone is genuinely the right answer.
A seed-stage SaaS on Team
$26 base, 12 GB of logs (7 chargeable GB at $0.50 is $3.50), 12 uptime monitors (11 chargeable at $1.00 is $11.00) and 8 cron monitors (7 chargeable at $0.78 is $5.46). Total $45.96 a month, $551.52 a year. Note that $19.96 of that monthly bill, 43% of it, is meters rather than the plan.
A Series A company on Business
$80 base, 40 GB of logs ($17.50), 20 GB of application metrics ($7.50), 35 uptime monitors ($34.00), 25 cron monitors ($18.72), 200 hours of UI profiling ($50.00) and a month of continuous profiling on one service ($23.00). Total $230.72 a month, $2,768.58 a year, of which $150.72 a month is metered usage. At that level the reserved-volume discount is no longer optional: at the 19.4% gap above, moving $150 of monthly pay-as-you-go onto reserved capacity saves about $29 a month, roughly $350 a year, for filling in a form.
The Seer AI add-on
Sentry sells an AI debugging agent called Seer as a separate subscription. Sentry's pricing page states that it requires its own subscription but does not print a rate in the served HTML, so we are not going to assert one. Third-party pricing trackers report roughly $40 a month per active code contributor, which would make a ten-engineer team about $400 a month, more than a Business subscription with meaningful overage. Treat that as an unverified third-party figure and get it in writing from Sentry before you budget for it.
How to keep a Sentry bill predictable
Four things move the number more than anything else, in rough order of impact.
Set sample rates deliberately. Traces, profiles and replays all default to sampling decisions that are fine at low traffic and expensive at high traffic. A traffic spike is exactly when you least want a surprise invoice.
Filter noisy errors at the SDK. Browser extension errors, bot traffic and known third-party script failures burn error quota and tell you nothing. Filtering them before they leave the client is free.
Reserve volume once your usage is stable. The roughly 20% gap between reserved and pay as you go is pure margin left on the table if your consumption is predictable.
Do not buy breadth by the monitor. Uptime and cron coverage are the two meters that scale with how careful you are rather than with how much traffic you have, which makes them the worst thing to meter. A per-monitor price is a standing incentive to leave the boring jobs unwatched, and the boring jobs, the nightly backup and the weekly export, are precisely the ones that fail silently. That is the whole argument for dead man's switch monitoring on a flat plan rather than a metered one.
One practical note if you are reconciling a Sentry invoice against what you actually provisioned: the usage numbers live in Sentry's API, the monitor inventory usually lives in your infrastructure code, and the invoice lives in your billing system. Getting all three into one place so the meters can be checked against reality is ordinary data-integration work between vendor APIs, and it is worth doing once rather than re-deriving it every renewal.
Is Sentry worth it?
For error monitoring and tracing, yes, and it is not close. Sentry is the reference product in that category, the SDK coverage is the best in the market, and $26 a month for 50,000 errors with 90 day retention is fair. The place to be careful is treating it as an all-in-one observability bill. The plan price is a small part of what most accounts actually pay, the Team to Business upgrade adds no volume, and the availability features are metered per monitor on a 60 second floor. Buy Sentry for what it is excellent at, then price the meters separately and decide each one on its own merits. The same discipline applies to Datadog's synthetics meter, which is priced on a comparable model and catches teams the same way.