BugSnag vs Raygun

BugSnag

Rest of world

Sovereignty: 1 of 6 dimensions proven

Full evaluation →

Raygun

Rest of world

Sovereignty: 1 of 6 dimensions proven

Full evaluation →

As of 6 Nov 2026, the panel rates BugSnag 3/5 and Raygun 2.5/5 in Error Tracking & APM. Weighted by what each judge cares about, 6 of the 6 judges pick BugSnag and 0 pick Raygun. BugSnag scores higher on 7 of 7 criteria (European sovereignty, Data scrubbing, retention & volume control, APM, tracing & profiling, plus 4 more). Provenance — BugSnag: Rest of world; Raygun: Rest of world, New Zealand. Ratings are the AI panel’s opinion, formed from public evidence: how scores are made.

The short answer

BugSnag leads the means on performance tracing (6.5 to 5.2; 6 lean BugSnag, 0 Raygun, 0 tie), data controls (6.8 to 4.8) and sovereignty (3.5 to 1.3), both leaning 6 to 0 with 0 ties, pricing transparency (5.2 to 4.2; 4 lean BugSnag, 2 tie), security and reliability (6.5 to 5.8; 3 lean BugSnag, 3 tie) and workflow and integrations (6.3 to 6.0; 2 lean BugSnag, 4 tie). Error tracking is the genuine split: 6.5 to 6.3, 2 judges lean BugSnag, 1 Raygun, 3 tie — the table's only Raygun lean — on resolve-in-version workflows and an audit log of grouping changes. All six sovereignty attributes — legal entity jurisdiction, data residency, hosting options, open source, data processing agreement, transfer impact assessment — read unknown for both products. Weighted totals span 6 to 4.5 for the skeptic judge, closest 6.1 to 5.4 for the SRE judge; the lean differs by judge and criterion.

Choose BugSnag if

  • You need OpenTelemetry-based tracing with documented sampling and quota behaviour across your stack — performance tracing leans BugSnag 6 judges to 0 (means 6.5 to 5.2).
  • Your compliance process requires retroactive server-side redaction of event data — data controls leans BugSnag 6 judges to 0 (means 6.8 to 4.8).
  • You need published prices before talking to sales — Free at $0/month with 7.5K events, 1M spans and 7-day retention, Select starting at $20/month and Preferred starting at $33/month are public, and events beyond quota are saved at no additional cost; pricing transparency leans BugSnag 4 judges to 0, with 2 ties.
  • Your team works in Jira and ships across mobile and backend platforms — a two-way Jira sync and versioned SDKs are documented, and workflow and integrations leans BugSnag 2 judges to 0, with 4 ties.
  • Your procurement rubric weights sovereignty — that criterion leans BugSnag 6 judges to 0, with means 3.5 to 1.3.

Choose Raygun if

  • You close errors at release time — Raygun's resolve-in-version workflows tied to deployment tracking drew the panel's only Raygun lean, in error tracking (1 judge leans Raygun, 2 lean BugSnag, 3 tie).
  • You need custom grouping you can audit — stack-trace-based grouping with custom grouping keys and an audit log of grouping changes were credited under error tracking (mean 6.3).
  • Your alerts must fire per error group — Raygun documents alerts settable per error group, part of its workflow and integrations score of 6.0, where 4 judges tied.
  • Your servers run.NET, Ruby or Node.js — Raygun's server-side APM covers those runtimes, with millisecond flame charts and N+1 detection rules documented.
  • You must contract outside the United States — the verdicts name SmartBear Software Inc. of Somerville, Massachusetts for BugSnag and Raygun Limited under New Zealand law for Raygun.

Read this comparison as one judge. Each weighs the same scores by what they care about.

The Data Protection Officer

Data protection officer of a statutory health insurer whose member portal and app are built by an in-house team and two agencies. Knows that a stack trace or a request body can carry a member number, a diagnosis code or an IBAN. Reads the subprocessor list, asks where events are stored and for how long, wants scrubbing on the server and not only in the SDK, asks which AI model explains the errors and where, and needs a DPA before the first SDK goes into production.

BugSnag

This judge's pick

Raygun

Criterion by criterion

Error capture, grouping & triage

BugSnag

Stack traces with source maps and symbol uploads, breadcrumbs, custom metadata, request and user fields, and an inbox prioritised by severity and user impact are documented, with named SDKs for Android, iOS, Flutter, React Native, Expo, Unity and JavaScript plus PHP/Laravel guides and snippets for Node, Java,.NET, Python, Go and Ruby. Issue states with automatic reopening, release and deploy reporting, and filtered views with bookmarks are also documented. I found no public information on custom fingerprinting, merge or split of issues, suspect commits, per-frame local variables or crash-free session and user counts.

Raygun

Automatic stack-trace grouping with a custom grouping key, message-based and provider-based variants, and an audit trail of grouping changes and merges is documented, together with a complete status workflow — active, resolved, ignored, permanently ignored — reactivation on new instances, resolve-in-version, deployment tracking, filtering by version, host, OS and browser, and affected-user tracking. I found no public information on breadcrumbs, mobile symbolication, crash-free sessions or suspect commits, and the SDK list is a general claim rather than a per-platform catalogue.

APM, tracing & profiling

BugSnag

Distributed tracing built on OpenTelemetry with gRPC and HTTP ingestion, collector support, head-based sampling in the SDK and tail-based rules via a collector, selectable P50 to P99 percentiles and trimmed means, CPU and memory system metrics, web vitals, and one-click navigation from an error to its trace are all documented. Sampling behaviour is unusually well explained, with the trace as the sampling unit, dynamic quota distribution and warnings when the error margin grows. I found no public information on code-level profiling, slow query or N+1 detection, background job monitoring, or host and container metrics beside the traces.

Raygun

Server-side traces are documented per millisecond with a call tree, flame chart and hot path, database and dependency calls captured with durations and credentials excluded, custom performance rules such as N+1 queries, sampling configurable per trace or per URL, and real user monitoring with Core Web Vitals. I found no public information on percentile timings per endpoint, distributed tracing across services, OpenTelemetry ingestion, or host and container metrics beside the traces.

Alerting, integrations & developer workflow

BugSnag

Notification rules for new, frequent, reopened and collaborator-actioned errors with basic and saved filtersets, daily and weekly digests, Slack, Microsoft Teams, Twilio and Datadog, and a Jira integration with two-way sync and automatic issue creation once a set number of events arrive in a time window are documented, alongside public APIs for data access, builds, error reporting and traces, and routing of alerts to code owners. An MCP server surfaces errors in the IDE with AI-powered fix suggestions, though the pages do not name which model runs them or where it runs — the first question I would ask. I found no public information on PagerDuty or Opsgenie routing, GitHub or GitLab issue tracker integrations, or a CLI.

Raygun

Twenty-five integrations are listed including Slack, PagerDuty, VictorOps, Jira OAuth, GitHub and GitLab plus webhooks, alerts can be configured per error group, new issues trigger email, and there is an APM daily digest and a Slack daily performance summary. AI fix suggestions are documented and the provider is named — OpenAI, available only on explicit opt-in — which is exactly the disclosure I look for. I found no public information on a documented REST API or two-way tracker sync, and the command line documented is for configuring the APM agent, not for releases and source maps.

Data scrubbing, retention & volume control

BugSnag

Server-side redaction rules over the user, request, device, breadcrumb and metadata roots, creatable from project settings or on the fly from a live event and applied retroactively, are the scrubbing I need before any member data touches an error payload; quota behaviour is equally clear, with events beyond quota saved at no additional cost, an automatic upgrade after three overages in thirty days, a rate-limiting option, spans never overage-charged, and a documented on-premises edition. Sensitive data management is gated to the Enterprise plan and capped at ten rules per project, and retention is stated only as seven days on the free tier. I found no public information on IP anonymisation or an SDK-side opt-out of request bodies and user data.

Raygun

Developers control what is sent, sensitive fields can be filtered out before sending in all providers, IP collection can be disabled per plan, data is removed from online storage and backups when retention expires or on manual deletion, identified users and their PII can be permanently deleted, and trace sampling is fully controllable per trace or per URL with discarded occurrences exempt from the quota. All documented scrubbing happens before events are sent; I found no public information on server-side scrubbing rules, retention durations per plan, or a self-hosted edition.

Access control, security & the vendor's reliability

BugSnag

SAML single sign-on ships on the Preferred plan and SSO user provisioning with sensitive data management on Enterprise, roles such as Organization Admin and Project Owner are named, encryption in transit and at rest, SOC 2 Type II, ISO 27001/27017/27018, CSA STAR, regular independent assessments and a responsible disclosure policy with a 72-hour confirmation and 90-day confidentiality are published, and the status page shows 90 days of 100% uptime per component. I found no public information on audit logs, a published SLA or service credits, or how the SDKs behave when the ingest endpoints are unavailable. The terms exclude downtime costs from recoverable damages and cap liability at twelve months of fees.

Raygun

SAML2 single sign-on is documented with named identity providers, organisation-wide enforcement and auto-provisioning, TOTP two-factor authentication, granular team permissions, TLS in transit and encrypted storage, third-party pentests every six months, and a public status page with monitored components. I found no public information on ISO 27001 or SOC 2 with stated scope, SCIM provisioning, a published SLA with credits, a vulnerability disclosure programme, or documented SDK behaviour when the vendor itself is down.

European sovereignty

BugSnag

The contracting entity is SmartBear Software Inc. of Somerville, Massachusetts, with an EU representative and European headquarters in Galway and Irish law for customers outside the US and Canada; the privacy policy references a Data Processing Addendum and states personal information is processed in the US and overseas under standard contractual clauses. I found no public information on where error events and traces are stored, on an EU data region, on a subprocessor list, or on where the AI features run. An on-premises edition is documented, but the captured pages do not state that every event stays on the customer's own servers.

Raygun

The contracting entity is Raygun Limited, a New Zealand company under New Zealand law, with US customers possibly paying Raygun Incorporated, hosting on Amazon Web Services with no region stated, and OpenAI named as the third-party AI provider. I found no public information on an EU data region, a data processing agreement, or a subprocessor list — without a DPA and a named chain, nothing of ours would go into production.

Pricing transparency

BugSnag

The pricing page publishes a free tier at $0 per month with 7.5K events, 1M spans and 7 days of retention, Select starting at $20 and Preferred starting at $33 per month, annual billing at a discount, a 14-day Preferred trial without a card, and a custom-priced Enterprise plan behind contact. Quota behaviour is unusually well documented: events beyond quota are saved at no additional cost, an automatic upgrade follows three overages in thirty days, minidump events count as five, and spans are never overage-charged. The volume the $20 and $33 starting prices cover is not stated, retention for paid plans is not stated, and beyond US sales tax the pages leave taxes to the terms, which exclude them from the fees.

Raygun

Public figures exist: on-demand errors at $0.001 per error, 60% savings on reserved events versus on-demand, a 33% annual discount, a 14-day free trial with no credit card required, and SAML SSO free on Business and Enterprise or $50/month as an add-on, billed in US dollars. I found no public information on the individual plan prices or retention pricing, and Enterprise is "Talk to us Custom", so a full annual cost cannot be computed from the captured pages.

Sovereignty, side by side

Dimension BugSnag Raygun
Legal entity Not determined Incorporated in NZ
Hosting location Not determined Not determined
Hosting options Not determined Not determined
Open source Not determined Not determined
Data processing agreement Published, self-serve Not determined
Transfer impact assessment Not determined Not determined

Facts, side by side

Only facts both products carry under the same definition — anything else would not be a fair row.

Fact BugSnag Raygun
Compliance · Certifications PCI DSS · ISO 27001 · ISO 27017 · ISO 27018 · SSAE 16 / ISAE 3402 Type II (SOC 1, SOC 2, SOC 3) · CSA STAR · EU GDPR · CCPA1

captured 6 Nov 2026 · Report an error

HIPAA · GDPR · CCPA · PCI2

captured 6 Nov 2026 · Report an error

Compliance · Encryption yes3

captured 6 Nov 2026 · Report an error

yes4

captured 6 Nov 2026 · Report an error

Data · Retention retained per stated uses and legal/financial/compliance obligations5

captured 6 Nov 2026 · Report an error

until retention period expires or manual deletion6

captured 6 Nov 2026 · Report an error

Legal · Entity SmartBear Software Inc. and affiliates5

captured 6 Nov 2026 · Report an error

Raygun Limited · 18996736

captured 6 Nov 2026 · Report an error

Legal · Entity name SmartBear Software Inc. · 450 Artisan Way, Somerville, MA 02145, USA7

captured 6 Nov 2026 · Report an error

Raygun Limited · 18996738

captured 6 Nov 2026 · Report an error

Legal · Governing law Ireland · Massachusetts7

captured 6 Nov 2026 · Report an error

New Zealand8

captured 6 Nov 2026 · Report an error

Legal · Terms last updated 2026-11-047

captured 6 Nov 2026 · Report an error

January 20248

captured 6 Nov 2026 · Report an error

Pricing · Annual discount yes9

captured 6 Nov 2026 · Report an error

3310

captured 6 Nov 2026 · Report an error

Pricing · Late payment interest 0.0157

captured 6 Nov 2026 · Report an error

1.5 · 148

captured 6 Nov 2026 · Report an error

Pricing · Trial all Preferred features · 14 · no9

captured 6 Nov 2026 · Report an error

14 · no11

captured 6 Nov 2026 · Report an error

Questions about BugSnag and Raygun

Which is better, BugSnag or Raygun?

As of 6 Nov 2026, the panel rates BugSnag higher: 3/5 against 2.5/5. Weighted by what each judge cares about, 6 of the 6 judges pick BugSnag and 0 pick Raygun. Which one fits depends on the criteria that matter to you; the table on this page shows both side by side, and each judge’s reading has its own page. Both were scored by the same AI bench under the same rubric, from public evidence only, as the methodology describes.

Is BugSnag or Raygun European?

BugSnag is not European by provenance (Rest of world). Raygun is not European by provenance (Rest of world, New Zealand). Provenance is read from each company’s own legal pages: who builds it and who owns it.

Do BugSnag and Raygun have a free plan?

As of 6 Nov 2026, BugSnag’s own pages state a free plan; Raygun’s pages as captured do not. Each product page links the pricing page it was read from.