Overview
Coroot takes a different route from SDK-based APM: an eBPF agent on each Linux node watches network traffic and processes at the kernel level, so it can map services, measure latency and errors for HTTP, databases and other protocols, and collect logs, traces and profiles without changing application code.
On top of that data it builds a service map, SLO tracking, deployment tracking, cloud cost monitoring and alerting, and an analysis engine (with AI in the paid edition) that walks dependencies to suggest the likely root cause of an incident.
It is always self-hosted, so telemetry never leaves your infrastructure. The Community Edition on GitHub (7,300+ stars) is free and fine for small clusters; Standard costs $1 per monitored CPU core a month, which makes costs depend on cluster size rather than data volume.
Pricing and plans
checked 28 Sept 2026Free tier · no card
Open-source Community Edition with the core features is free to self-host; paid editions have a 14-day trial.
| Plan | Price | What you get |
|---|---|---|
| Community Edition | $0 | Open source on GitHub, core features, self-hosted |
| Standard | $1/monitored CPU core/mo | eBPF monitoring, SLOs, alerting, cost monitoring, continuous profiling, AI root cause analysis, SSO, RBAC, business-hours support |
| Premium | Custom | 24x7 and phone support, premium onboarding, air-gapped installs, training |
What it does
- eBPF auto-instrumentation
- Collects metrics, logs, traces and profiles with zero code changes.
- Service map
- Discovers services and dependencies automatically from network traffic.
- SLO tracking and alerting
- SLOs and alerts sent to Slack, PagerDuty or Teams.
- Continuous profiling
- CPU and memory profiles linked to services.
- Root cause analysis
- Dependency-aware analysis, AI-assisted in the Standard edition.
- Cost monitoring
- Cloud cost per service and node.
Strengths and limitations
Strengths
- No code changes needed to get service-level metrics and traces.
- Free open-source edition.
- Predictable per-core pricing that does not grow with data volume.
- Data stays inside your own infrastructure.
Limitations
- Self-hosted only; there is no managed cloud.
- eBPF agent requires Linux nodes.
- AI root-cause analysis, SSO and RBAC need the paid Standard edition.
- No frontend, uptime or error-grouping features.
Who it suits
Good for
- Kubernetes clusters with services you cannot instrument easily
- Teams that must keep telemetry on-premises
- Platform teams wanting fast service maps and SLOs
Look elsewhere if
- Teams wanting a hosted SaaS
- Serverless apps with no nodes to run an agent on
Alternatives to Coroot
| Tool | From | Free option | Stages |
|---|---|---|---|
| Amazon CloudWatch AWS-native metrics, logs, alarms and tracing | Usage-based | Free tier | |
| BugSnag Error and stability monitoring for mobile and web | $20/mo | Free tier | |
| Checkly Synthetic and uptime monitoring as code | $24/mo | Free tier | |
| Grafana Cloud Hosted metrics, logs, traces and dashboards | $19/mo | Free tier | |
| Honeycomb Tracing and event-based observability | $150/mo | Free tier | |
| New Relic Full-stack observability platform | $10/mo | Free tier |
Questions
Is Coroot free?
Yes, the Community Edition is open source and free to self-host; the Standard edition costs $1 per monitored CPU core a month.
Does Coroot need code changes?
No, it uses eBPF to collect telemetry at the kernel level without instrumenting code.
Is there a hosted Coroot?
No, Coroot is self-hosted so your data stays in your infrastructure.
How is Coroot priced?
Per monitored CPU core: for example, 40 cores cost $40 a month on Standard.
Sources read for this page