Sentry pricing

About

Sentry is a developer-first application monitoring platform that combines error tracking, distributed tracing, logging, session replay, uptime monitoring, and profiling in a single product suite, alongside Seer — an AI debugging agent that performs root-cause analysis and opens fix pull requests directly against a connected repository. The company positions itself around fixing issues “without compromising on velocity,” and its own site states it serves 200,000+ organizations and 4 million+ developers across 100+ platforms, processing more than 190 billion events per day, with named customers including GitHub, Disney, Autodesk, Anthropic, Atlassian, and Duolingo.

Sentry sells directly to individual developers through a self-serve free tier plus monthly or annual paid plans, and to larger engineering organizations through a sales-led Enterprise tier with a dedicated technical account manager and custom quotas. Pricing is built around a base subscription (Developer/Team/Business) that bundles a fixed monthly quota per data category, with pay-as-you-go rates covering usage above that quota — positioning Sentry as a hybrid subscription-plus-usage vendor rather than a pure per-seat or pure-metered one, competing with observability incumbents (Datadog, New Relic) and open-source/self-hosted alternatives (GlitchTip) on developer experience and granular, product-by-product metering.

Pricing summary : How Sentry’s tiered-base-plus-usage pricing model works

Sentry uses a hybrid tiered-base-plus-usage model with four dimensions:

1. Base plan tier: Free Developer (1 user), Team ($26/mo annual, $29/mo monthly), Business ($80/mo annual, $89/mo monthly) — both with unlimited users — and custom-quoted Enterprise.
2. Bundled monthly quota: Every paid plan includes the same base allotment regardless of tier — 50K errors, 5GB logs, 5GB application metrics, 5M spans, 50 replays, 1 cron monitor, 1 uptime monitor, 1GB attachments, and 100 size-analysis builds.
3. Pay-as-you-go overage: Usage above the bundled quota draws from a shared PAYG budget billed per data category, volume-tiered and cheaper per unit at higher volume — e.g. errors range from $0.0003625 (Team, 50K–100K band) down to $0.0001500 (Team, 20M+ band); logs and application metrics are a flat $0.50/GB; cron monitors are $0.78 each and uptime monitors $1.00 each; UI profiling is $0.25/hour and continuous profiling $0.0315/hour. Business PAYG rates run roughly 3x Team’s for errors ($0.0011125 vs $0.0003625 in the entry band) and 2x for spans ($0.0000040 vs $0.0000020), while logs, metrics, replays, monitors, profiling, and attachments are priced identically across Team and Business.
4. Seer AI add-on: A separate charge of $40/active contributor/month (any user opening 2+ PRs to a Seer-connected repo that month), layered on top of any paid plan.

What makes this different: Instead of pricing by seat count, Sentry keeps users unlimited on paid plans and instead meters nearly every telemetry category independently (errors, spans, logs, metrics, replays, monitors, profiling hours, attachments) through one shared PAYG budget — so the bill scales with data volume across products, not headcount.

Pricing by product

Sentry Monitoring Platform (Individual plans)

Sentry Monitoring Platform (Business plans)

Sales motions across products: PLG / self-serve for Developer, Team, and Business (credit-card signup, 14-day trial); sales-led for Enterprise (Contact Sales).

Per-event-type usage rates (pay-as-you-go, above bundled quota)

All rates read from docs.sentry.io/pricing/ (accessed 2026-09-01); Sentry notes its published per-category prices are rounded.

Seer AI debugging agent (add-on)

Hidden costs : What real Sentry teams actually pay once every meter stacks

The advertised $26 Team / $80 Business headline understates what real teams pay once errors, spans, logs, replays, monitors, and Seer AI all draw from the same pay-as-you-go budget. Two real-world examples:

Growing Team-plan startup running Seer AI

The $26 sticker is a rounding error next to the Seer AI line alone — once a team turns Seer on for even a handful of active contributors, the AI add-on outweighs every telemetry-overage line combined, and the stacked per-event-type overages (errors, spans, logs, replays, monitors) each add a few dollars that compound into real money. This is the classic multi-meter bill-shock pattern: no single line is scary, but ten of them stacked together are.

Business-plan team crossing the new post-2025 span line

Before August 27, 2025, this same 8M-span/month account paid only the errors and logs overage ($114) because 8M spans still fit inside the old 10M included quota — the span-quota cut alone added the $12 tracing line with zero change in actual usage, pushing the effective bill up roughly 10% overnight.

Want to estimate your own Sentry bill? Use the Sentry pricing calculator to model your monthly cost based on error volume, spans, logs, replays, and Seer AI contributor count.

Pricing evolution : From flat transactions to a five-year run of meter rewrites

Cadence

Tracked range: 2022-01–2026-09. 2022 Q1–Q4 is verified stable via 7 bracketing Wayback snapshots. The 2023-04–2024-06 and 2024-10–2025-08 windows were not independently captured via Wayback; the events dated inside them above are sourced directly from Sentry’s own docs, changelog, blog, and Help Center articles rather than screenshots — see Sources.

Notable changes

• 2023-02-13 — Session Replay reaches general availability as a new billed product line.
• 2023-05-09 — Performance Monitoring moves to unit-based “Performance Units” pricing alongside the Profiling launch.
• 2024-06-11 — Performance Units retired; Tracing rebilled directly in spans, and Session Replay’s included quota is cut from 500 to 50 replays/month — the single largest mechanic overhaul in Sentry’s pricing history.
• 2024-09-12 — A second migration wave completes moving remaining legacy accounts onto spans-based billing.
• 2025-08-27 — Team/Business included span quota is cut from 10M to 5M spans/month, with the per-span overage rate unchanged.
• 2025-09-03 — Sentry Logs reaches general availability as a new metered SKU.
• 2026-01-27 — Seer AI is re-priced from $20/month + $25 usage credits to a flat $40/active-contributor/month.

The June 2024 pricing overhaul in detail

The June 11, 2024 cutover is Sentry’s biggest documented pricing migration: it retired the Performance Units model introduced barely a year earlier (May 2023) in favor of billing distributed tracing directly in spans, and simultaneously cut the Session Replay allotment bundled with every paid plan by 90% (500 → 50 replays/month). Sentry’s own docs site still hosts a dedicated “before June 11, 2024” legacy-pricing page rather than deleting the old model outright, and a further Help Center article dated September 12, 2024 describes a second wave moving remaining legacy accounts across — evidence that the migration ran in stages over several months rather than as a single flip.

What’s unique : The mechanics that separate Sentry from seat-priced observability rivals

1. Users are never metered — telemetry categories are. Every paid Sentry plan includes unlimited users; instead of taxing headcount, Sentry runs roughly ten independent per-event pay-as-you-go meters (errors, spans, logs, application metrics, replays, cron monitors, uptime monitors, UI profiling, continuous profiling, attachments) off one shared spend budget, so a five-person team generating heavy telemetry pays more than a fifty-person team generating light telemetry — the inverse of most observability-platform pricing built around per-seat or per-host fees.

2. Seer AI bills by contributor activity, not seats or tokens. At $40 per “active contributor” per month — any user who opens 2+ pull requests to a Seer-connected repo that month — Sentry chose a usage signal tied to actual shipped code rather than metering AI usage directly in tokens, scans, or fixes (as its own legacy Seer pricing briefly did). That’s a distinctive middle path between flat per-seat AI add-ons and pure-metered AI billing, and one worth watching alongside the broader shift from entitlements to AI usage credits.

3. The base-plan sticker price is frozen while the metering underneath keeps changing. Team ($26/$29) and Business ($80/$89) have not moved since at least January 2022, yet the metering model underneath has been rebuilt three separate times in the same window — transactions, then Performance Units, then spans — with Session Replay, Logs, and Seer AI all layered in as additive metered SKUs rather than forcing customers into new plan tiers.

Strengths & weaknesses

Billing UX : Named controls on Sentry’s pricing and cost-estimation surfaces

• Payment Frequency toggle (Monthly / Annual) — switches the pricing-page plan cards between monthly and annual-equivalent rates; defaults to Annual, which shows the lower $26/$80 headline prices versus $29/$89 under Monthly.
• “Estimate your monthly cost” calculator — an embedded per-category slider tool (errors, logs, application metrics, replays, spans, cron/uptime monitors, attachments, profiling hours, Seer) that computes an estimated total bill in real time as each usage slider moves, with a per-line-item cost breakdown shown alongside the total.
• Paid Plan Tier / Payment Frequency selectors inside the calculator — let a prospect model Team vs. Business costs and Annual vs. Monthly billing before signing up, without leaving the pricing page.
• “A closer look at each plan” comparison grid — a full feature-by-feature table (Developer/Team/Business/Enterprise columns) spanning monitoring, workflow & alerts, reporting & search, integrations, account & data management, admin & compliance, and support.
• Spend notifications / maximum spend threshold — an account-level control (visible in the feature comparison grid) that lets Team, Business, and Enterprise customers cap PAYG spend and get notified as usage approaches it.

Strategic wins : Why Sentry’s monetization decisions have held up

1. Keeping users free while metering telemetry captures the real cost driver

Most SaaS pricing anchors on seats because it’s easy to explain, but for an observability platform the marginal cost driver is data volume, not headcount. Sentry’s decision to make users unlimited on every paid plan while metering roughly ten separate telemetry categories directly ties price to the usage metric that actually matters, rather than taxing collaboration the way per-seat rivals do.

2. A flat, contributor-based price for Seer avoids per-token sticker shock

Rather than metering Seer’s AI usage in tokens or per-scan credits (as its own legacy pricing briefly did), Sentry moved to a flat $40/active-contributor/month fee — a deliberate hedge against the unpredictable variable-cost problem that plagues token-metered AI add-ons, at the cost of some revenue upside on its heaviest AI users.

3. Spend notifications and a maximum-spend threshold give buyers a self-serve safety net

Letting Team, Business, and Enterprise customers cap PAYG spend and get notified as usage approaches it is a rare self-serve control most usage-based vendors reserve for enterprise contracts, directly addressing the threshold-and-alerting gap that drives most usage-pricing complaints.

4. A public per-category rate card kept pricing legible through three model rewrites

Even as the underlying meter changed from transactions to Performance Units to spans, Sentry has kept every rate publicly listed on docs.sentry.io/pricing rather than moving to sales-only quotes — unusual discipline for a company that rebuilt its meter three times in four years.

Areas to improve : Where Sentry’s own pricing mechanics create friction

1. The August 2025 span-quota cut was never explained in a blog post or changelog entry

Two Help Center articles record the change, but neither Sentry’s blog nor its public changelog carries an equivalent announcement. For a change that doubled tracing-overage risk for a meaningful slice of customers, a comparable-visibility explanation — in the spirit of transparent usage-based billing communication — would reduce the surprise-bill complaints visible on Trustpilot and in third-party teardowns.

2. Ten-plus independent meters make total-cost forecasting genuinely hard

Errors, spans, logs, metrics, replays, cron monitors, uptime monitors, UI profiling, continuous profiling, attachments, and size-analysis builds each carry their own quota and rate — useful for granularity, but it multiplies the entitlement and usage-grant complexity a finance team has to model. A single blended cost-simulator export (CSV or API), alongside the existing slider calculator, would help FinOps teams reconcile actual invoices against the estimate.

3. Business’s PAYG rate premium over Team is not surfaced on the plan cards

A prospect comparing the $26 vs. $80 base prices has no visibility that Business’s overage rate for errors and spans runs roughly 2-3x Team’s rate for the identical unit until they open the calculator. A one-line footnote on the plan cards themselves would close this gap before a customer signs up expecting Team-level overage math.

4. Three metering-model rewrites in four years leave stale numbers scattered across the web

Third-party pricing guides and comparison sites frequently cite pre-2024 quotas or rates from the transactions/Performance-Units era. A persistent “pricing model version” banner — as Sentry’s own legacy-pricing doc already carries for the June 2024 cutover — applied site-wide would reduce confusion for buyers researching via search.

Key takeaways

1. A frozen base price can coexist with a constantly-evolving usage layer. Sentry hasn’t touched its $26/$80 sticker since at least 2022, yet it has rebuilt its usage-metering model three times in the same window — proof that price stability and pricing-model evolution are separable decisions.
2. New product lines are easiest to introduce as new meters, not new plans. Session Replay, Logs, and Seer AI all launched as additive metered SKUs layered onto the existing Team/Business/Enterprise structure rather than forcing customers into new plan tiers — see our guide on tracking and metering usage events for the underlying mechanics.
3. Quota reductions are price increases, even when the sticker price doesn’t move. The August 2025 span-quota halving raised real bills for customers in the 5M-10M/month band without changing a single published rate — a reminder that usage-based pricing has more levers than the price-per-unit number alone.
4. Flat per-active-user AI pricing is a legitimate alternative to metering AI usage directly. Seer’s move from credit-based Issue Scan/Fix pricing to a flat $40/active-contributor fee shows a viable middle path between pure per-token billing and pure per-seat billing for an AI feature.
5. Multi-meter granularity is a double-edged design choice. It lets customers pay only for the telemetry categories they actually use, but it also multiplies the number of places a bill can surprise a customer — every additional independent meter is both a precision win and a bill-shock risk, as documented in our piece on unpredictable AI and usage costs.

UBP implications

1. Observability is a natural home for per-category metering because telemetry volume is genuinely heterogeneous. Errors, spans, logs, and replays have wildly different cost-to-serve profiles, so blending them into one generic “event” unit sacrifices pricing accuracy that a multi-meter model like Sentry’s captures — directly relevant to the per-event pricing billing-unit pattern.
2. Hybrid pricing’s “predictability half” can anchor for years while the “usage half” is rebuilt repeatedly. Sentry is a clean natural experiment: the fixed fee is untouched since 2022, but the usage layer has had three structural rewrites — a model other hybrid-pricing vendors could deliberately replicate to change monetization mechanics without triggering the sticker-shock of a headline price change.
3. Quota cuts deserve the same disclosure rigor as price increases. Sentry’s August 2025 span-quota halving shows that reducing what’s included is functionally a price increase for affected accounts, yet it shipped via Help Center articles rather than a blog post — a governance gap that usage-based pricing teams elsewhere should close proactively rather than reactively.

Sources

• Sentry Pricing page (accessed 2026-09-01)
• Sentry Pricing & Billing docs (accessed 2026-09-01)
• Sentry Pricing docs — Quotas (accessed 2026-09-01)
• Sentry for Enterprise (accessed 2026-09-01)
• Sentry Legacy Pricing docs (before June 11, 2024) (accessed 2026-09-01)
• Sentry Changelog: Introducing Performance Units (accessed 2026-09-01)
• Sentry Blog: Introducing Session Replay (accessed 2026-09-01)
• Sentry Blog: Seer — debug with AI at every stage of development (accessed 2026-09-01)
• Sentry Blog: Logs in Sentry — Now in Open Beta (accessed 2026-09-01)
• Sentry Help Center: How is my plan changing? (August 27, 2025) (accessed 2026-09-01)
• Sentry Help Center: What is the pricing for Seer? (January 21, 2026) (accessed 2026-09-01)

Bottom line

Sentry’s pricing is a study in decoupling: the $26/$80 base fee hasn’t moved since at least 2022, but the usage engine underneath it has been rebuilt three times — transactions, then Performance Units, then spans — while new product lines (Session Replay, Logs, Seer AI) arrive as additive meters rather than forced upgrades. That granularity buys real per-team accuracy, but it also means the bill can move in a dozen different places at once, most visibly in the August 2025 span-quota cut that raised real costs without touching a single published rate.

Want to compare Sentry against other observability and developer-tools pricing? Browse the pricing blueprint, or see how Datadog and New Relic structure their own usage-based observability pricing.

Pricing timeline : Major events on a vertical axis

Each milestone below corresponds to a public pricing change, product launch, or material adjustment. Major events use a filled marker; minor adjustments use a faded one.

Current pricing capture

Sep 2026

Blueprint capture of Sentry's current pricing: a free Developer tier, Team (~$26/mo annual / $29/mo monthly) and Business (~$80/mo annual / $89/mo monthly) base plans bundling a fixed monthly quota, granular pay-as-you-go overages across errors, spans, logs, application metrics, replays, cron/uptime monitors, profiling hours, and attachments, and Seer AI priced separately at $40/active contributor/month.

Seer AI re-priced to flat $40/active contributor

Jan 2026

Sentry's blog (blog.sentry.io/seer-debug-with-ai-at-every-stage-of-development/, published 2026-01-27) announced Seer moved from its original launch pricing ($20/month subscription plus $25 of usage credits metering Issue Scans and Issue Fixes individually) to a flat $40-per-active-contributor-per-month price for unlimited use, where an active contributor is any user opening 2+ PRs to a Seer-connected repo that month. Pre-2026 subscribers were grandfathered onto the old 'Legacy Seer Pricing.'

Sentry Logs reaches general availability

Sep 2025

Sentry announced general availability of its Logs product on 2025-09-03 (press release via businesswire.com/Sentry, and blog.sentry.io/logs-in-sentry-open-beta/ for the preceding open beta), adding a new metered SKU — 5GB of logs bundled per plan, billed at $0.50/GB beyond that — alongside the existing errors/spans/replays meters.

Team/Business span quota cut from 10M to 5M

Aug 2025

Two Sentry Help Center articles dated 2025-08-27 ('How is my plan changing?' and 'How is my Team plan changing?', sentry.zendesk.com) confirm the included monthly span quota on Team and Business was halved from 10M to 5M spans, with the per-span pay-as-you-go rate unchanged — so accounts sending 5M-10M spans/month went from zero tracing overage to paying for every span above the new, lower line. Third-party cost guides (e.g. urgentry.com) estimate the added annual cost at roughly $100-1,000+ depending on volume.

Second migration wave for remaining legacy-plan customers

Sep 2024

A follow-up Sentry Help Center article ('How is my legacy plan changing? (September 12, 2024)', sentry.zendesk.com) describes a further migration wave moving remaining legacy transactions/Performance-Units accounts onto the spans-based model roughly three months after the June 2024 cutover.

Performance Units retired for span-based Tracing

Jun 2024

Sentry's own docs mark this as the hard cutover date (docs.sentry.io/pricing/legacy-pricing/, 'Pricing & Billing before June 11, 2024'): the Performance Units/transactions meter was retired for new and upgrading accounts in favor of billing Tracing directly in spans (confirmed live in the 2024-06 Wayback capture: 10M spans included on Team/Business). The same snapshot shows Session Replay's included allotment cut from 500 to 50 replays/month, and Metrics (beta) and LLM Monitoring (beta) appear as new feature rows — the single biggest mechanic overhaul in Sentry's pricing history, though the $26/$80 base price itself did not move.

Performance Monitoring becomes 'Performance Units'

May 2023

Per Sentry's official changelog (sentry.io/changelog/2023-5-9-introducing-performance-units), the May 9, 2023 launch of Profiling moved Performance Monitoring to a unit-based meter: a plain transaction consumes 1 Performance Unit, a transaction with Profiling attached consumes 1.3 Performance Units — replacing the flat reserved-transactions quota with a profiling-aware multiplier.

Session Replay added as a metered product

Apr 2023

Session Replay went GA on 2023-02-13 (blog.sentry.io/2023/02/16/introducing-session-replay-from-sentry-bridge-the-gap-between-code-and-ux) with a 2-billing-cycle free extension for beta users. By the 2023-04-23 Wayback snapshot it appears as a new billed dimension on the pricing calculator (500 replays included, reserved up to 1M), alongside a new Codecov add-on ($29/mo for the first 5 seats). Errors/Transactions quotas were unchanged from 2022; the reserved transaction ceiling had grown to 100M by 2023-02.

Baseline: Errors + Transactions model

Jan 2022

Earliest captured Wayback snapshot of sentry.io/pricing: Developer (free), Team ($26/mo annual), Business ($80/mo annual), and custom Enterprise, all metering 'Errors' and 'Transactions' (Performance Monitoring) as the two usage dimensions, plus Attachments. Team/Business included 50K errors (up to a 10M reserved ceiling) and 100K transactions (up to 30M), with pay-as-you-go for anything above.