What is Sentry?
Sentry is an application monitoring and debugging platform made by Functional Software, Inc., which does business as Sentry. SDKs installed in an application capture unhandled errors, performance spans, logs and, where enabled, session replays and profiles, and send them to a Sentry server. There, error events are grouped into issues with stack traces, breadcrumbs, tags and release information, and traces connect a request across frontend, backend and other services. The about page states that Sentry is used by over 4 million developers. The LICENSE file carries a copyright notice starting in 2008.
The server in the getsentry/sentry repository is a Python application built on Django, with a TypeScript and React frontend; the current pyproject.toml requires Python 3.13 or later. Sentry is offered in two forms. The hosted service at sentry.io stores data in one of two regions chosen when an organization is created, the United States (Iowa) or the European Union (Frankfurt), and the data storage location cannot be changed afterwards. The second form is self-hosted Sentry, a Docker Compose setup maintained in the getsentry/self-hosted repository.
According to the self-hosted documentation, the self-hosted package is a minimal setup for simple use cases without guarantees or dedicated support. It requires Docker 19.03.6 and Docker Compose 2.32.2 or later, 4 CPU cores, 16 GB of RAM plus 16 GB of swap (32 GB of RAM recommended) and 20 GB of free disk, and on x86-64 hosts ClickHouse 26.6 and later need the x86-64-v3 feature set with AVX2. The docs describe it as the Business plan without software limitations, but billing, spike protection, data storage location selection and Seer and other AI and ML features, which are closed source, are not part of it.
Sentry uses calendar versioning. The latest GitHub release is 26.10.1, published on October 28, 2026. Its notes list new UI components such as a split panel and menu, persisted query transformations in Dashboards and Explore saved queries, a low-priority query control for dashboards, new Inbox item models, a delete action for custom integrations and changes to Investigations and AI conversation views.
Sentry is not released under an OSI open source license. The code is published under the Functional Source License 1.1 with an Apache 2.0 future license (FSL-1.1-Apache-2.0): it may be used for any purpose other than a Competing Use, meaning a commercial product or service that substitutes for Sentry, and each version becomes available under Apache License 2.0 two years after it is released. Sentry describes this model as Fair Source. The SDKs are maintained in separate repositories. Similar projects in this directory include PostHog, Highlight.io and SigNoz.
Key capabilities
- Issues and error grouping Unhandled errors and captured exceptions are grouped into issues that can be filtered, triaged, assigned, resolved or archived. An issue page shows an event count chart, tag distribution such as browser, release, URL and environment, the stack trace with source context, highlights and an activity log. Issue tracking links to GitHub, Linear and other trackers are available from the issue sidebar.
- SDKs for many platforms The README lists official SDKs for JavaScript, Electron, React Native, Python, Ruby, PHP, Laravel, Go, Rust, Java and Kotlin, Objective-C and Swift, C# and F#, C and C++, Dart and Flutter, Perl, Clojure, Elixir, Unity, Unreal Engine, Godot Engine and PowerShell. Setup guides per framework are in the platform documentation.
- Tracing and spans Distributed tracing follows a request across frontend, backend and other services. The trace view shows a waterfall of spans with durations, Web Vitals such as LCP, FCP, INP, CLS and TTFB for page loads, and span attributes in a side panel. The Developer, Team and Business plans each include 5M spans.
- Explore and trace queries Explore is a query interface for spans, logs, metrics and errors. Users choose an aggregate such as count, group and sort results, view span samples, trace samples or aggregates, and save or compare queries. Release 26.10.1 persists transformations in saved queries.
- Logs Structured logs sent through the SDKs are stored next to errors and traces and can be searched by attributes such as severity, release, trace and custom fields, as described in the logs documentation. Each plan includes 5 GB of logs, with additional volume at 0.50 USD per GB.
- Session Replay Session Replay plays back web and mobile user sessions around an error, with breadcrumbs, console, network, errors, trace, memory and tags tabs. Text and media can be masked, and the replay view counts dead clicks and rage clicks. Paid and free plans include 50 replays.
- Profiling Continuous profiling captures CPU profiles to show slow functions in production code, as described in the profiling documentation. Profiling is listed in the Debug group of the product documentation next to traces and logs.
- Insights for frontend, backend, mobile and AI agents Insights pages summarize transactions per minute, p50 and p75 durations, failure rates, slow outgoing HTTP domains and render-blocking assets per project. An Agents view traces agent workflows, LLM calls, tool executions and MCP servers.
- Cron and uptime monitoring Cron monitors check that scheduled jobs run on time, and uptime monitors check URLs at a fixed interval. Each plan includes one cron monitor and one uptime monitor, with additional cron monitors at 0.78 USD each and uptime alerts at 1.00 USD each, according to the pricing page.
- Alerts and metric monitors Monitors and alerts notify on issue conditions, metric thresholds, cron failures and uptime checks, with routing to tools such as Slack and PagerDuty. The Developer and Team plans include 20 metric monitors and Business includes 1,000.
- Dashboards and releases Custom and prebuilt dashboards chart errors, spans and other data; Developer allows 10 custom dashboards, Team 20 and Business unlimited. Releases track crash-free sessions, version adoption and failure rates per deploy.
- User Feedback User Feedback collects bug reports from end users and links them to the related errors, replays and tags.
- Seer Seer is Sentry's AI feature for root cause analysis, fix generation through Autofix and questions about telemetry, documented under AI in Sentry. It is not available on the Developer plan, requires a subscription at additional cost on Team, Business and Enterprise, and is not part of self-hosted Sentry. The pricing page announces credit-based Seer pricing from December 8.
- Mobile build tools Size Analysis monitors mobile app build sizes for regressions, Snapshots catch visual changes on pull requests, and Build Distribution delivers app builds to internal testers.
- REST API and integrations The Sentry API manages organizations, projects, issues, events, releases and other resources; on sentry.io, region-specific domains us.sentry.io and de.sentry.io are used. API access and third-party integrations are listed as Team plan features, and the Developer plan lists MCP access.
- Self-hosted Sentry The self-hosted setup is installed by cloning the self-hosted repository at a release tag, running
./install.shand starting the stack withdocker compose up --wait; the web interface listens on port 9000. iOS symbolication is not built in, Android system symbols cannot be redistributed, and PlayStation support requires permission from Sony. - Plans on sentry.io Developer is free for one user with 5k errors and a 30-day lookback. Team costs 29 USD per month, or 26 USD per month billed annually, and Business costs 89 USD per month, or 80 USD per month billed annually, both with unlimited users, 50k errors and up to 90-day lookback. Business adds code owners, anomaly detection, SAML and SCIM, advanced inbound filtering, a BAA and Relay; Enterprise is priced on request.
Where Sentry fits
Tracking production errors to a code change
An SDK reports an exception with its stack trace, breadcrumbs, release and environment. Sentry groups it into an issue, shows when it was first and last seen and in which release, and links the stack frames to source files when code mappings to a repository are configured. Issues can then be assigned, linked to a tracker and resolved in a release.
Diagnosing slow pages and endpoints
With tracing enabled, each page load or request becomes a trace whose spans show browser, resource, database and HTTP timings. The trace waterfall marks Web Vitals for page loads, and Explore aggregates span durations by operation or transaction to find slow endpoints, slow queries or render-blocking assets.
Reproducing user-reported problems
Session Replay records the DOM and user interactions around an error, so a developer can watch the steps that led to it next to console output, network requests and the related trace. User Feedback reports submitted from the application are attached to the same replay and error where available.
Monitoring scheduled jobs and endpoints
Cron monitors track whether scheduled jobs run as expected and can trigger alerts through the monitors and alerts system. Uptime monitors request a URL at a fixed interval and show availability over time per project and environment.
Correlating logs with traces and errors
Logs sent through the SDK carry the trace and span IDs of the code that emitted them. In Sentry, a log line can be opened from its trace, and logs can be filtered by severity, release or custom attributes alongside the errors from the same request.
Running Sentry on own infrastructure
Organizations that keep monitoring data on their own servers use the self-hosted Docker Compose setup, which the documentation sizes at 4 CPU cores and 16 GB of RAM plus swap as a minimum. The project recommends regular upgrades through the release tags, and AI features such as Seer and the data storage location choice remain specific to sentry.io.
Reviews of Sentry
Be the first to share your experience with Sentry.
Write a review
Loading review form…
Found Sentry useful? Upvote it todayYour vote helps other IT teams find it in the daily ranking.




