Pay per bug, not per crash
Every plan is priced by new issues — distinct bugs — per month. A bad deploy that throws the same error two million times counts once. All plans work with the official Sentry SDKs: you only change the DSN.
简单透明的价格
Solo
€90/年 — 免费 2 个月
适用于自由职业者和独立开发者。
- 500 个新问题/月 · 100k 事件上限
- 30 天保留期
- 3 名团队成员
- 邮件 + webhook 提醒
- AI agent API
Team
€290/年 — 免费 2 个月
适合产品团队和成长型 SaaS
- 每月 5,000 个新问题 · 75 万事件上限
- 90 天数据保留
- 无限成员
- Email + webhook + Slack、Discord、Telegram、Teams
- AI agent API + 作用域令牌
- 回归检测
- 发布健康度 (crash-free)
- Status page + 自动分配
Business
€1 290/年 — 免费 2 个月
适合大规模成熟产品
- 每月 25,000 个新问题 · 300 万事件上限
- 180 天数据保留
- 无限成员
- 所有告警渠道
- 优先支持
- 50 个监控器 · 2,500 次 AI 分析
- 发布健康度 (crash-free)
- 5 个 status pages + 自动分配
Scale
€2 990/年 — 免费 2 个月
适用于高流量产品和平台。
- 100,000 个新问题/月 · 1500 万事件上限
- 180 天数据保留
- 无限成员
- 所有告警渠道
- 优先支持 + 入门指导
- 200 个监控 · 10,000 次 AI 分析
- 发布健康度 (crash-free)
- 20 个 status pages + 自动分配
所有套餐均包含完整 Sentry SDK 兼容性、EU 托管、GDPR 合规和 AI agent API 访问。无设置费用。随时取消。
Compare plans
Every number in this table is the limit the product actually enforces.
Swipe sideways to see all plans →
| 功能 | Free | Solo | Team | Business | Scale |
|---|---|---|---|---|---|
| Price per month | €0 | €9 | €29 | €129 | €299 |
| Price per year | — | €90 | €290 | €1 290 | €2 990 |
| New issues per month (billed unit) | 50 | 500 | 5 000 | 25 000 | 100 000 |
| Event ceiling per month | 10 000 | 100 000 | 750 000 | 3 000 000 | 15 000 000 |
| Event retention | 14 天 | 30 天 | 90 天 | 180 天 | 180 天 |
| Longest retention with add-on | — | 90 天 | 270 天 | 360 天 | 360 天 |
| Team members | 1 名团队成员 | 3 名团队成员 | 无限成员 | 无限成员 | 无限成员 |
| Uptime & heartbeat monitors | 1 | 3 | 10 | 50 | 200 |
| AI analyses per month | 5 | 50 | 500 | 2 500 | 10 000 |
| Public status pages | — | — | 1 | 5 | 20 |
| Webhook alerts | Not included | Included | Included | Included | Included |
| Slack, Discord, Telegram & Teams alerts | Not included | Not included | Included | Included | Included |
| Regression detection | Not included | Not included | Included | Included | Included |
| Release health (crash-free rate) | Not included | Not included | Included | Included | Included |
| Spike protection | Not included | Not included | Included | Included | Included |
| Quiet hours & escalation | Not included | Not included | Included | Included | Included |
| Ownership rules & auto-assign | Not included | Not included | Included | Included | Included |
| Teams with project visibility | Not included | Not included | Included | Included | Included |
| Unused quota rolls over | Not included | Included | Included | Included | Included |
On every plan, Free included
Official Sentry SDKs · email alerts · source maps · AI agent API and MCP server · EU hosting · data scrubbing with IP removal on by default
What counts toward your plan
One unit you pay for, one safety net — and no surprises when you hit either.
New issues — the unit you pay for
An issue is one distinct bug. Events with the same fingerprint — the exception type plus your in-app stack frames, or a fingerprint you set yourself — are grouped into one issue. An issue counts once, in the calendar month (UTC) when it first appears. Repeats of a known bug, a resolved bug that comes back, and new events of older issues cost nothing extra.
Deleting or merging issues doesn't give quota back — the monthly counter only goes up.
Event ceiling — a safety net
Each plan also has a monthly ceiling on raw events. It is deliberately generous: it exists so that a runaway loop can't flood the storage, not to bill you per event.
One quota per account
All projects in an account share the plan's quota. Adding a project doesn't add quota, and the count of a deleted project stays with the account for that month.
At the limit we don't go blind
Past the quota, issues keep being grouped, counted and alerted on — you still see what breaks and how often. Only the full detail of individual events stops being stored. Far above the quota, at a multiple of it, ingestion pauses until the next month. Nothing is charged unless you have turned on overage yourself.
Unused quota carries over
From the Solo plan up, what you don't use in a month is added to the next one — from the previous month only, so you never have more than double your plan.
Spike protection
From the Team plan up, a sudden flood of events can be sampled: counts and alerts stay exact, but only a share of raw events is stored, so one bad deploy can't use up the month's ceiling.
Paid add-ons
Both are optional, switched on per account in the console and invoiced monthly together with your plan.
更长时间保留事件
为整个账户延长保留期，随时可取消。
| 套餐 | Included | ×2 | ×3 |
|---|---|---|---|
| Solo | 30 天 | 60 天 +€4,50 per month | 90 天 +€9 per month |
| Team | 90 天 | 180 天 +€14,50 per month | 270 天 +€29 per month |
| Business | 180 天 | 360 天 +€64,50 per month | — |
| Scale | 180 天 | 360 天 +€149,50 per month | — |
The Free plan has no add-ons. A dash means the step would exceed the storage limit, so we don't sell it.
保留期上限为 365 天——我们不出售存储无法保留的天数。
超出配额时
By default, past your quota we keep counting issues but stop storing event detail. Instead, you can let new issues through for €1 per 100 extra new issues.
- •默认关闭——不开启就不会产生任何费用。
- •月度上限由你设定。超过后我们又会丢弃，账单不会超出预期。
- •按新错误计费，与套餐相同的单位，而不是按事件计费。
Example: 250 new issues over the quota are 3 started blocks of 100 → €3 that month.
附加功能手动开票——不会自动扣款。
OpenErrorAPI vs Sentry
Sentry is an excellent and much broader product. If error tracking is what you need, here is an honest side-by-side — including where Sentry is the better choice.
| 功能 | OpenErrorAPI | Sentry |
|---|---|---|
| What you pay for | New issues — distinct bugs. Repeats of a known bug are free; raw events only have a generous ceiling. | Error events — every occurrence counts toward your volume. |
| Free plan | 50 new issues · 10 000 event ceiling · 14-day retention · 1 名团队成员 | 5 000 errors · 30-day lookback · 1 user |
| Paid plans start at | €9/mo (Solo) · Team €29/mo | $26/mo billed yearly or $29 monthly (Team) |
| Volume on the Team plan | 5 000 new issues + 750 000 event ceiling per month · 无限成员 | 50 000 errors per month · unlimited users |
| Retention | 14–180 days by plan, up to 360 days with the add-on | 30 days on the free plan, up to 90 days on Team and Business |
| Over the quota | Issues keep being counted and alerted on; only event detail pauses. Paid overage is opt-in, with a monthly cap you set. | Events over your quota are not accepted; extra volume comes from reserved volume or a pay-as-you-go budget. |
| Where your data lives | EU servers we run ourselves — accounts and events alike.* | US (Iowa) or EU (Frankfurt), chosen once when the organization is created; accounts, tokens and audit logs always stay in the US. |
| SDKs | The official Sentry SDKs — change only the DSN | The official Sentry SDKs |
| Product scope | Error tracking, release health, uptime & heartbeat monitors, status pages, user feedback | All of that, plus tracing, session replay, profiling, logs and more |
| AI & agents | AI root-cause analysis and search (monthly allowance, or unlimited with your own key); MCP server and scoped agent API on every plan | Seer AI debugging agent as a separate subscription (Team and up); MCP access |
* Optional AI features send the relevant error context to Anthropic (USA) — or to your own AI provider if you add your own API key. Privacy Policy
Sentry figures as published on sentry.io and docs.sentry.io, checked on 2026-11-02. Sentry prices are in USD and can change — check their current pricing before you decide. [1] [2] [3]
When Sentry is the better choice
- •You need performance tracing, session replay or profiling — we deliberately don't build those.
- •Your mobile apps need native symbolication (dSYM, ProGuard).
- •You want to self-host the whole stack.
- •Procurement requires a SOC 2 report or SAML/SCIM single sign-on.
- •You want one vendor for errors, performance, logs and replays.
When OpenErrorAPI is
- •Error tracking is what you actually use, and you want to pay for distinct bugs, not traffic.
- •Your error volume is spiky — a bad deploy shouldn't cost money or blind you for the rest of the month.
- •You want all data, account data included, on EU servers.
- •You want AI agents to triage errors through a scoped API or MCP, without giving them admin rights.
Frequently asked questions
What exactly is a new issue?
One distinct bug, counted once — in the month when we see it for the first time. The ten-thousandth occurrence of the same exception is free. Events are grouped by fingerprint: the exception type plus your in-app stack frames, or a fingerprint your SDK sends.
Do I need a credit card?
No. New accounts start with a 14-day trial of the Team plan and then continue on Free automatically. Paid plans are switched on at your request from the console and invoiced — nothing is charged automatically.
Is paying yearly cheaper?
Yes. A year costs 10 monthly payments, so 2 months are free.
What happens if I go over my plan?
We keep grouping, counting and alerting; only event detail pauses until the next month or an upgrade. Far above the quota, ingestion stops for the rest of the month. Paid overage is off unless you turn it on and set a monthly cap.
Can I upgrade, downgrade or cancel at any time?
Yes. When you move to a smaller plan, alert channels, monitors and rules you already set up keep working; the new limits apply only to what you add afterwards.
Do AI features cost extra?
Each plan includes a monthly number of AI analyses (see the table). If you add your own Anthropic or OpenAI API key, that limit doesn't apply — you pay your provider directly. The AI agent API and the MCP server are on every plan.
Where is my data stored?
On servers in the EU that we run ourselves: MySQL for accounts and issues, ClickHouse for events. Passwords, tokens and similar fields are scrubbed from events before they are stored, and IP addresses are removed by default. Read the Privacy Policy
Do my existing Sentry SDKs work?
Yes. Keep the official Sentry SDK and replace the DSN. The documentation lists exactly what is supported and what isn't. See the compatibility matrix
What happens to my data when retention ends?
Events older than your retention are no longer available. Issues — title, counts, first and last seen — stay until you delete them, so your history doesn't vanish with the events.