Error monitoring at this size is not a discretionary purchase. You ship several times a day across dozens of services and clients you do not control; without a de-duplicating error tracker attached to releases, a regression that affects 3% of sessions is invisible until a customer writes in. Sentry's paid plans include unlimited users, so it is one of the very few lines on your renewals list where hiring does not raise the bill.
That is exactly why the plan price misleads people. $26 a month for Team and $80 for Business are rounding errors on an engineering budget — $312 and $960 a year. The bill an org this size actually receives is the data: errors, logs, spans, replays and profiles above the included volumes. And Sentry's own published rate card shows Business charging two to three times the Team rate for every one of those errors.
So the honest framing of the upgrade — usually driven by wanting SAML and SCIM, which sit on Business — is not "$54 a month for SSO". It is: at our current event volume, what does moving every event onto the Business rate card cost us for the year? Work that number out before the renewal call. It is the only version of this decision that is about money rather than a feature checkbox.
The money, named
What you pay for the plan. Sentry publishes Team at $26 a month and Business at $80 a month, both when billed annually with the default pre-paid volume, and both with unlimited users. Enterprise is custom-priced, so it is absent from this page. Every paid plan includes the same monthly allowance: 50,000 errors, 5 GB of logs, 5 GB of application metrics, 5 million spans, 50 replays, 1 uptime monitor, 1 cron monitor, 100 size-analysis builds and 1 GB of attachments.
What you pay for the data. Above those allowances Sentry publishes a per-event rate card, tiered by volume, with a lower rate for volume you reserve in advance and a higher one for pay-as-you-go. The errors table is the one that matters most at this size:
| Error volume band | Team, reserved | Business, reserved | Business ÷ Team |
|---|---|---|---|
| >50k – 100k | $0.00029 / error | $0.00089 / error | 3.1x |
| >100k – 500k | $0.000175 / error | $0.0005 / error | 2.9x |
| >500k – 10M | $0.00015 / error | $0.0003 / error | 2.0x |
| >10M – 20M | $0.00013 / error | $0.00026 / error | 2.0x |
| >20M | $0.00012 / error | $0.00024 / error | 2.0x |
Put a real volume through it. Two million errors a month is unremarkable for a company of this size with mobile clients and a chatty third-party SDK or two. Applying the published bands:
TEAM, 2,000,000 errors / month
50,000 in the >50k-100k band x $0.00029 = $14.50
400,000 in the >100k-500k band x $0.000175 = $70.00
1,500,000 in the >500k-10M band x $0.00015 = $225.00
plan fee = $26.00
total = $335.50 / month = $4,026 / year
BUSINESS, same 2,000,000 errors / month
50,000 x $0.00089 = $44.50
400,000 x $0.0005 = $200.00
1,500,000 x $0.0003 = $450.00
plan fee = $80.00
total = $774.50 / month = $9,294 / year
The Business premium at this volume = $5,268 / yearNow do it at ten million errors a month, which is what a bad quarter or one unfiltered bot looks like: Team is $1,535.50 a month, or $18,426 a year; Business is $3,174.50, or $38,094. The premium is $19,668 a year — for the same errors, the same retention and the same alerting, differing in features like SAML, SCIM, unlimited dashboards and advanced quota management.
| Errors per month | Team, per year | Business, per year | The premium |
|---|---|---|---|
| 50k (included) | $312 | $960 | $648 |
| 500k | $1,326 | $3,894 | $2,568 |
| 2M | $4,026 | $9,294 | $5,268 |
| 10M | $18,426 | $38,094 | $19,668 |
And the largest line may not be errors at all. Sentry publishes its Seer AI debugging agent at $40 per active contributor per month, on top of Team, Business or Enterprise. That is the only per-person meter in the product, and it is not a seat: Sentry counts an active contributor as any user who opens two or more pull requests in a month to a Seer-enabled repository, reset monthly — so the number to multiply is your committing engineers, not your headcount. It scales with your engineering org rather than your traffic: 100 active contributors is $48,000 a year, 250 is $120,000, 500 is $240,000 — our arithmetic on the published rate. Whatever you conclude about Team versus Business, decide Seer separately, and decide it per team rather than per org.
What it has to beat. On the Team plan at two million errors a month, the whole thing costs $4,026 a year, which at a $180,000 fully-loaded cost per engineer and 2,000 hours a year is 45 engineer-hours — under an hour a week across the entire org. For a tool that turns "a customer says checkout is broken" into a stack trace with a release attached, that is not a close call. The argument worth having is about the rate card, not about the purchase.
The reason, at this size
1. The SSO upgrade is metered, and that is genuinely unusual.
Plenty of vendors charge more for a plan that includes SAML. Sentry's published rate card does something different: moving to Business also re-prices every error you send, at two to three times the Team rate, for the whole year. That makes "we need SSO for the security review" a decision whose cost depends on your traffic, which is not how anybody instinctively evaluates a feature gate.
None of that is hidden — Sentry publishes both rate cards openly, which is more than most of this list does — but you have to multiply to see it. Do the multiplication with your volume before the renewal, then ask whether the security requirement can be met another way (network controls, an SSO-enforcing browser policy, restricting who has an account at all) at less than the number you just calculated. Sometimes it can, sometimes it cannot; either way you will be arguing about $19,668 rather than $54.
2. Event volume is an engineering hygiene variable, not a fixed cost.
Unlike seats, your event count is something your own code decides. A single noisy client-side integration, an unfiltered crawler, a retry loop that reports every attempt, or a deprecated endpoint still being hit by an old mobile build can add millions of events a month that nobody will ever read.
Sentry's own quota-management documentation exists precisely because of this: inbound filters, per-project rate limits, spike protection, sampling and better fingerprinting are all published controls. The half-day somebody spends on them is worth $1,509 a month at the Team rate and $3,094 at the Business rate in the ten-million-error example above — and, more importantly, it makes the issue stream readable again. Cost control and signal quality point the same direction here, which is rare.
3. Reserved volume is cheaper than pay-as-you-go, and it expires.
Sentry publishes two prices for every category: a lower reserved rate you pay for in advance, and a higher pay-as-you-go rate for anything above it, drawn from a shared budget. The documentation also states plainly that unused reserved volume expires at the end of each billing month, and that data sent after both the reserved volume and the PAYG budget are exhausted is dropped and not charged for.
That is a two-sided trap. Reserve too much and you burn the difference every month; reserve too little and either your PAYG budget absorbs it at the higher rate or you silently stop being monitored for the rest of the cycle. The right posture at this size is to reserve close to your median month and keep a PAYG budget sized for a bad one — and to alert on approaching the ceiling, because losing error monitoring mid-incident is a far worse outcome than an overage.
4. Unlimited users is a real structural advantage — use it.
Both paid plans include unlimited users. In a stack where nearly every other line is per-seat, that means the cheapest thing you can do with Sentry is give everyone access: support engineers, QA, the on-call rota, product managers investigating a report. Restricting Sentry accounts to save money is one of the few false economies where the vendor is not even charging you for the thing you are rationing.
The exception, and it is an important one, is Seer at $40 per active contributor per month. That meter is per person, so "turn it on for everyone" and "turn Sentry on for everyone" are opposite decisions financially. Keep them in separate conversations.
5. Five separate meters share one budget, and only errors are obvious.
Errors are what people think Sentry bills for. The published allowances cover five more categories — logs, application metrics, spans, replays and attachments — each with its own rate card, and logs and metrics are published at $0.50 per GB above the included 5 GB on both paid plans. Turning on tracing across a large service estate, or session replay across a high-traffic front end, can change your bill more than any plan decision on this page.
So budget per category, not per product, and roll each new category out to one service or one surface first with the real volume measured before it goes estate-wide. "We turned on replay last sprint" should never be the explanation for a bill anybody had to investigate.
What to do instead of just upgrading
Four controls. About a day of one engineer's time, and they decide a four- to five-figure annual line at this size.
1. Price the Business upgrade at your actual event volume.
Take last quarter's monthly error count, run it through both rate cards band by band exactly as above, and put the annual difference in the renewal document. That figure — not $54 a month — is what the SAML requirement costs. Do the same for logs, spans and replays if you send meaningful volume in those categories.
2. Spend half a day on inbound filters and fingerprinting first.
Bots, browser-extension noise, health-check failures, one endpoint reporting every retry: cut those before you buy more volume, because they are also why nobody reads the issue stream. At the published rates, a million junk errors a month is $150 a month on Team and $300 on Business.
3. Reserve for the median month, budget PAYG for the bad one, alert on both.
Unused reserved volume expires monthly and exhausted budgets drop data — Sentry documents both. Size reserved volume at your median, keep a pay-as-you-go budget for spikes, and set an alert well before the ceiling. Losing error monitoring during the incident that caused the spike is the worst possible way to save money.
4. Decide Seer per team, with a review date.
At $40 per active contributor per month it is the only meter here that scales with headcount: $48,000 a year at 100 contributors, $240,000 at 500. Run it with two or three teams for a quarter, agree in advance what "it worked" looks like, and expand or stop on evidence rather than on enthusiasm.
What would change our mind
- You have a hard, documented requirement for SAML or SCIM. Then Business is not optional and the argument stops being about the rate card — but you should still know, to the dollar, what the upgrade costs at your volume before signing, and you should still cut junk events first.
- Your error volume is genuinely tiny. Under the included 50,000 a month, the whole decision is $312 versus $960 a year and none of the arithmetic here is worth a meeting.
- You are already paying for an APM platform that does error tracking to the same standard. Then this becomes a consolidation question rather than a purchase question, and the right comparison is against the incumbent's own published rates.
- Sentry changes the rate card or the included volumes. Every figure here is a published per-event or per-GB rate. If the bands, the reserved/PAYG split or the 50,000-error allowance change, the arithmetic on this page has to be re-run from the new tables.
- You can show us a negotiated Enterprise rate. We price at published list because that is what is verifiable; Sentry publishes no Enterprise price and we do not estimate one. If you have a signed rate, use yours — the volume-hygiene argument holds either way.
What we checked
Every figure above came from a page we opened, so you can check our work before you forward this to finance.
- Sentry — Pricing — Developer free (one user); Team $26 a month and Business $80 a month, both when billed annually with default pre-paid data, both with unlimited users; Enterprise custom; Seer AI debugging agent $40 per active contributor per month as an add-on to Team, Business or Enterprise; logs and application metrics +$0.50 per GB above the included 5 GB. Read 30 August 2026.
- Sentry Docs — Pricing & Billing — Included monthly volume on every paid plan (50k errors, 5 GB logs, 5 GB application metrics, 5M spans, 50 replays, 1 uptime monitor, 1 cron monitor, 100 size-analysis builds, 1 GB attachments); the reserved-versus-pay-as-you-go model; the statement that unused reserved volume expires at the end of each billing month and that data beyond reserved volume and PAYG budget is dropped and not charged; and the per-error rate table used above (Team reserved $0.00029 / $0.000175 / $0.00015 / $0.00013 / $0.00012 and Business reserved $0.00089 / $0.0005 / $0.0003 / $0.00026 / $0.00024 across the published volume bands, with Sentry noting the listed prices are rounded). Read 30 August 2026.
- Sentry Docs — Quota management — The published controls for managing volume — inbound filters, rate limits, spike protection and sampling — referenced in the controls section. Read 30 August 2026.
Provenance. Every plan price, add-on price, included volume and per-event rate on this page was read from a Sentry page on 30 August 2026 — the plan prices and the Seer rate from sentry.io/pricing, the rate card and billing mechanics from Sentry's own pricing documentation. The annual totals, the band-by-band worked examples at 500k, 2M and 10M errors a month, the Business-÷-Team multiples, the Seer totals at 100, 250 and 500 contributors and the 45-hour break-even (at a $180,000 fully-loaded cost per engineer and 2,000 hours a year) are our arithmetic on those published figures, and are labelled as ours in each caption. Enterprise pricing is absent because Sentry publishes none. No error volume is assumed for your org: the volumes above are named examples chosen to show the shape of the meter, and you should substitute your own. No sponsorship, no affiliate link, no vendor review of this page. Last reviewed 30 August 2026. Corrections: [email protected].
Questions we get asked about this one
Is Sentry worth it for a 100–1,000 person engineering team?
Yes. Paid plans include unlimited users, so the tool does not get more expensive as you hire, and at two million errors a month the Team plan totals about $4,026 a year on Sentry's published rates — roughly 45 engineer-hours at a $180,000 fully-loaded cost. The decision that carries real money is not whether to buy it but whether to move to Business: that plan re-prices every error at two to three times the Team rate, which is $5,268 a year at two million errors a month and $19,668 at ten million.
What does Sentry cost at two million errors a month?
Applying Sentry's published reserved rates band by band: on Team, 50,000 errors at $0.00029, 400,000 at $0.000175 and 1,500,000 at $0.00015 comes to $309.50 a month, plus the $26 plan fee — $335.50 a month, or $4,026 a year. On Business the same volume is $694.50 plus the $80 plan fee, or $9,294 a year. Both figures are our arithmetic on Sentry's published rate card, and both exclude logs, spans, replays and attachments, which are metered separately.
Is Sentry Business worth it over Team?
Only if you can name what it buys and the number survives the multiplication. Business adds SAML and SCIM, unlimited custom dashboards and advanced quota management for $54 more a month — but it also charges two to three times the Team per-error rate on every event above the included 50,000. At two million errors a month that is $5,268 a year; at ten million, $19,668. If the driver is an SSO requirement from a security review, price it at your real volume first and check whether the requirement can be met another way for less.
How much is Sentry's Seer AI add-on?
Sentry publishes Seer at $40 per active contributor per month, added to a Team, Business or Enterprise plan. An active contributor is any user who makes two or more pull requests in a month to a Seer-enabled repository, counted fresh each month — a seat count is not the multiplier. It is the only per-person meter in the product, so it scales with the engineering org rather than with traffic: our arithmetic puts it at $48,000 a year for 100 active contributors, $120,000 for 250 and $240,000 for 500. At this size it can easily be the largest Sentry line you have, which is why it deserves its own trial, its own success criteria and its own decision per team.
How do we cut a Sentry bill without losing coverage?
In this order. Cut junk volume first — bots, extension noise, health-check failures and retry loops — using Sentry's published inbound filters, rate limits and sampling; at a million junk errors a month that is $150 on Team or $300 on Business every month, and it makes the issue stream readable. Then check whether you are on Business for a feature nobody uses, because the plan is charging a multiple on every event. Then size reserved volume at your median month with a pay-as-you-go budget for spikes, since unused reserved volume expires and exhausted budgets drop data. Then review the newer categories — logs, spans, replays — one surface at a time. None of those removes a single real error from view.
Tomorrow morning, five tools you haven’t heard of yet.
This verdict is free and always will be. The newsletter is the part you pay for: five new AI tools every weekday morning, one line on what each does, one line on why it matters for a PM, and the link. Start with this morning’s issue, free.
$10/month or $96/year after 7 free days, cancel in one click · a real Adoption Brief, in full · every issue, published free · [email protected]
Other verdicts at this size
- Is Atlassian Jira worth it for a 100–1,000 person engineering team? — Yes, on Standard — the tool is not the decision at this size, the edition is. Jira Cloud Standard is $36,500 a year at 500 users and $68,500 at 1,000, and the Premium step is $26,000 and $42,000 a year for a checkbox that applies to everyone on the site.
- Is LaunchDarkly worth it for a 100–1,000 person engineering team? — Yes, once you have counted service connections and client-side MAU — neither of which is your headcount. $10 per connection per month and $8.33 per 1,000 MAU means one extra environment across 80 services costs $9,600 a year, and the governance tier is quote-only.
- Is New Relic worth it for a 100–1,000 person engineering team? — Yes, if the user-type policy is written before the contract. Full platform users are $349 a month at Pro list against $49 for a core user, so fifty misassigned people cost $180,000 a year — several times the $0.40-per-GB data line everyone actually negotiates.
- Is Okta worth it for a 100–1,000 person engineering team? — Yes — identity is a control, not a convenience, at this size. The money is the suite: Okta publishes $6, $14 and $17 per user per month, so 500 identities is $36,000, $84,000 or $102,000 a year, and the upper suites only pay if somebody staffs the automation.
- Is PagerDuty worth it for a 100–1,000 person engineering team? — Yes, on Professional and billed annually. At 500 responders that is $126,000 a year against $246,000 on Business — and the money leaks through licences held by people who left the rota, at $252 each per year.
- Is Cursor worth it for a 100–1,000 person engineering team? — Yes, on Standard seats billed annually. At 500 engineers that is $192,000 a year against $576,000 for the all-Premium version of the same purchase — and the rework exposure is 4.7x to 38x the licence.
- Is Datadog worth it for a 100–1,000 person engineering team? — Yes, but only on a host count you measured yourself. Infrastructure Pro and APM on the same 500 hosts is $276,000 a year at list — and Datadog's own published retention rate says bills like yours grow about 20% a year.
- Is GitHub Copilot worth it for a 100–1,000 person engineering team? — Yes, on Business at $19 a seat. At 500 engineers that is $114,000 a year against a rework exposure of $900,000 to $7.4m — which is why the three controls matter more than the purchase order.
- Is Snowflake worth it for a 100–1,000 person engineering team? — Yes, on Standard at $2.00 a credit unless somebody can name the Enterprise feature. One X-Large warehouse left running is $280,320 a year — the settings decide this bill, not the contract.
- Should we buy an LLM observability platform? — The next observability purchase you will be pitched, and our call on it.
- Should we buy CodeRabbit? — Another developer-tool line sold per contributor, costed the same way.
- The do-not-adopt library — The calls we make against tools we think you should skip entirely.
Every published verdict is listed on the verdicts index.