Perses
CNCF dashboarding tool with dashboards-as-code, static validation, and Kubernetes-native workflows. A GitOps-friendly Grafana alternative.
| License | Apache-2.0 |
| Repository | https://github.com/perses/perses |
| GitHub stars (approx) | 2200 |
| Language | Go / TypeScript |
| Self-host deployment | docker, kubernetes helm, operator |
| Commercial hosted option | no / self-host only |
| Freshness status | Fresh |
| Freshness checked | 2026-06-23 |
| Categories | metrics, logs, traces, profiling |
| Replaces SaaS | grafana, datadog |
Perses is a dashboard tool and open dashboard specification for observability data, with support for Prometheus metrics, Tempo traces, Loki logs, and Pyroscope profiles. Its strongest angle is treating dashboards as validated resources that can live in Git and CI/CD instead of only in a UI database.
Pick it if: you want dashboard-as-code workflows, Kubernetes-native provisioning, embeddable panels, and an Apache-2.0 visualization layer for Prometheus-family stacks.
Skip it if: you need Grafana’s enormous plugin catalog, mature alerting surface, and long tail of community dashboards today. Perses is promising, but Grafana is still the compatibility default.
Deployment: Run it in a Docker or Podman container for small installs, use the Helm chart for Kubernetes, or add the Perses Operator when dashboards and datasources should be managed as cluster resources. You still bring the data stores.