The best long-term remote storage
for Prometheus

start storing and analyzing your metrics

in the most efficient way

Get Started


Simplifies monitoring

Offload non-trivial operational tasks for metrics storage such as replication, backup and seamless scalability to VictoriaMetrics. Just change a single line in Prometheus config

Global querying view

Query all the metrics your Prometheus instances collect via a single Prometheus-compatible datasource

Designed to be fast

Fast query engine. It excels on heavy queries over thousands of metrics with millions of metric values

Native PromQL support

VictoriaMetrics supports native PromQL. There is no need in learning yet another query language

Long Term Storage

VictoriaMetrics supports wide range of retention periods - from 1 month to 5 years

Common Table Expressions

Improve and simplify your queries by using WITH templates and extended PromQL



TimescaleDB and InfluxDB

See how VictoriaMetrics performs comparing to TimescaleDB and InfluxDB.

The tests were made via Time Series Benchmark Suite (TSBS) with following dataset: 4,000 simulated devices generated 10 CPU metrics every 10 seconds for 3 days (~100M reading intervals, ~1B metrics).

The benchmark was run in Google Compute Engine on two n1-standard-8 instances with 8 virtual CPUs, 30GB RAM and 200GB HDD.

When size matters High-cardinality
histogram image
  • 1.7Mingested metric values per second
  • 390MBcompact storage data size
  • 20xfaster query performance


Grafana image

Grafana playground

Grafana playground uses Prometheus-compatible datasource for VictoriaMetrics. The datasource supports native PromQL plus extended PromQL. Explore dasboards and try editing queries

Grafana playground
CTE image

Simplify complex PromQL with CTE

Feel the power of CTE for PromQL at `WITH` playground

WITH playground

Demo site

Play with fully-featured VictoriaMetrics at demo site


Single-server VictoriaMetrics

Try free single-server VictoriaMetrics

Precompiled binaries Docker images