Sentry

Sentry's Developer plan is free forever with 5,000 errors/month, 10,000 performance units, 50 session replays, and 1 team member. This is sufficient for personal projects and small applications. The Team plan starts at $26/month for 50,000 errors, 100,000 performance units, and 500 replays — typically covers early-stage startups. Business at $80/month adds more volume, SSO, custom dashboards, and SLA support. Self-hosted Sentry (open-source) is free with no limits but requires maintaining your own infrastructure.

Sentry is primarily used for: (1) Error tracking — capturing and grouping JavaScript, Python, mobile, and backend exceptions with full stack traces; (2) Performance monitoring — identifying slow database queries, API calls, and Core Web Vitals regressions; (3) Session replay — watching video-like recordings of user sessions that led to errors; (4) Release tracking — correlating error spikes with specific deployments; (5) Alerting — notifying teams via Slack, PagerDuty, or email when error rates spike. It's the default error tracking tool for web startups.

Use Sentry if you primarily need application error tracking and developer debugging tools — Sentry's error grouping, source maps, suspect commits, and session replay are purpose-built for debugging and have no equal. Use Datadog if you need infrastructure monitoring alongside application monitoring — Datadog covers server metrics, Kubernetes, network, logs, and traces in one platform. Many teams use both: Sentry for application errors (developer-facing) and Datadog for infrastructure observability (ops-facing). Sentry's free tier makes it the default starting point for most development teams.