Sentry Introduces Application Metrics: High-Cardinality, Trace-Connected Metrics for Application Debugging
Developers can capture application telemetry once and query it any way they need, with a direct path from metric spike to root cause.
SAN FRANCISCO — May 5, 2026 — Sentry, the leader in application monitoring, today announced the general availability of Application Metrics. With this launch, Sentry’s observability suite is now comprehensive, giving engineering teams errors, traces, logs, and metrics in one place. Application Metrics is high-cardinality and trace-connected, built for application debugging.
Traditional metrics tools require developers to define their questions before they instrument. Want to track checkout latency by region? That’s a new metric for every region — checkout_latency_usa, checkout_latency_canada, checkout_latency_mexico. By plan type? A new series for every plan and region. By browser? Start over again. Every new question means new instrumentation, a new deployment, and a wait to collect data, often long after an incident has already passed. Developers end up perpetually behind their own data, unable to ask what matters most in the moment it matters most.
Sentry Application Metrics changes that dynamic entirely. A developer instruments checkout latency once and attaches whatever context is relevant — region, plan type, browser, product category, user ID. From that point forward, any combination of those dimensions can be queried on demand, with no predefined aggregations and no re-instrumentation required. When a new question emerges mid-incident, the data is already there. This is made possible by Sentry’s decision to store metric events in full rather than pre-aggregating them.
Most metrics tools identify that a problem occurred. Sentry Application Metrics help developers answer why. Every metric event retains a direct connection to the trace it was emitted from, meaning a spike in any chart is one click away from exemplar traces, logs, and errors connected by trace ID. With Sentry Metrics, there is no pivoting between tools, no manual correlation by timestamp, and no guesswork.
“Developers don’t just want to know that something broke, they want to know why. For years, the tools haven’t matched that desire. Developers would see a spike in a chart and hit a wall when investigating because that spike didn’t provide the right context to understand what happened. They would spend unnecessary time correlating data from different tools in an effort to fill in the missing context. Application Metrics is Sentry’s answer to that wall. We want every developer to have the ability to go from ‘something is wrong’ to ‘here is exactly what happened and why.’ One tool, full context, no wasted time,” said Alex Jillard, Senior Engineering Manager at Sentry.
Getting started requires no additional infrastructure. If you’re already using Sentry, you can add metrics with your existing Sentry SDK. Instrument the signals you care about and they appear alongside your existing errors, traces, and logs.
Sentry Application Metrics is available now. For more information, visit sentry.io/product/application-metrics.
About Sentry
Sentry helps every developer detect, understand, and fix broken code, fast. Using Sentry’s debugging platform decreases resolution time from days to minutes, giving developers more time to do the stuff they love, all while making customers happier. Sentry is used by over 4 million developers and 150,000 organizations, including Disney+, Cloudflare, GitHub, Anthropic, Vercel, and Atlassian.
Contact
Cori Kendrick
Sentry
media@sentry.io