What it does
Grafana Alloy is an observability collector that combines pipelines for Prometheus, OpenTelemetry, Loki, and Pyroscope signals. Its configuration language connects components that read data, transform it, and send it to a chosen backend. Teams can use it as a single agent for several telemetry types instead of maintaining separate collectors for each signal. Alloy is software to install and configure, not a hosted storage or dashboard service.
Build a pipeline, not a dashboard
Install Alloy for the platform you use, then write a UTF-8 .alloy configuration describing the components in the data path. A component might read a log file, receive OTLP, collect metrics, or forward data to another service. Alloy evaluates the dependencies between those components, so a configuration can keep collection, processing, and export together. The getting-started guide introduces blocks, attributes, expressions, and formatting with alloy fmt.
For a first task, choose one source and one destination, configure those components, and run the file with alloy run path-to-config. The CLI reference says this command runs in the foreground and exposes an HTTP page for debugging Alloy and its components. Add additional signals only after the first pipeline is delivering the data you expect. This is an example workflow from the documentation; Alloy was not installed or run for this listing.
Alloy can be installed on Linux, Windows, macOS, FreeBSD, Docker, or Kubernetes, with architecture support varying by operating system. The software is Apache-2.0 licensed. It can collect and route telemetry, but a separate backend is needed to store and query it; Grafana Cloud fleet management is a separate hosted capability.
Grafana Alloy supports OpenTelemetry, but it is not the same product as the separately listed OpenTelemetry Collector. Alloy adds Grafana's component library and configuration language, including Prometheus, Loki, and Pyroscope pipelines.
Strengths
- Combines several telemetry pipelines in one collector, including Prometheus, OpenTelemetry, Loki, and Pyroscope components.
- A single configuration file can describe collection, processing, and export, with a CLI page for debugging component health.
- The collector is published as an Apache-2.0 open-source project.
Watch out for
- Users must install the collector and configure its components and destinations; Alloy does not itself provide a telemetry database or hosted dashboard.
- A broad pipeline can become complex to maintain, so teams should start with a small configuration and add components deliberately.
Best for
- Teams consolidating collection of metrics, logs, traces, and profiles
- Operators who want to configure a local or cluster agent rather than send all collection work to a hosted service
- Grafana ecosystem users who need Prometheus and OpenTelemetry pipelines in one agent
Sources
- Grafana Alloy documentation (Oct 25, 2026)
- Get started with Grafana Alloy (Oct 25, 2026)
- Install Grafana Alloy (Oct 25, 2026)
- run | Grafana Alloy documentation (Oct 25, 2026)
- grafana/alloy (Oct 25, 2026)
- Apache License, Version 2.0 (Oct 25, 2026)