---
title: "Introduction to Application Observability | Grafana Cloud documentation"
description: "What Application Observability powered by the knowledge graph is, how it differs from classic Application Observability, and key capabilities."
---

> For a curated documentation index, see [llms.txt](/llms.txt). For the complete documentation index, see [llms-full.txt](/llms-full.txt).

# Introduction to Application Observability

Application Observability powered by the knowledge graph is an application performance monitoring (APM) solution that combines OpenTelemetry-native instrumentation with the Grafana Cloud Knowledge Graph to provide deep visibility into your applications.

## What is Application Observability?

Application Observability helps software engineers reduce the time it takes to detect anomalies, identify root causes, and fix problems. It automatically discovers your services from telemetry data, surfaces performance insights, and connects metrics, logs, traces, and profiles in a unified view.

### Key capabilities

- **OpenTelemetry native**: Use open standards to collect metrics, logs, and traces without vendor lock-in.
- **Intelligent insights**: The knowledge graph automatically detects saturation, anomalies, failures, and errors without custom queries or alert rules.
- **RED metrics at a glance**: View latency, error ratio, and request rate for every service in the catalog.
- **Instrumentation quality**: Assess how well each service is instrumented and get actionable recommendations.
- **Integrated signal exploration**: Drill from a service into its metrics, logs, traces, and profiles without context switching.
- **Root cause analysis**: Use RCA workbench and Entity Graph to investigate correlated issues across your stack.
- **Service discovery**: The knowledge graph automatically discovers services and their relationships from your telemetry data.

## How it differs from classic Application Observability

Application Observability powered by the knowledge graph replaces the classic Application Observability experience for new Grafana Cloud organizations. Key differences:

Expand table

|                     | Classic Application Observability      | Application Observability (KG)              |
|---------------------|----------------------------------------|---------------------------------------------|
| Architecture        | Standalone plugin with Scenes-based UI | Built on the knowledge graph entity catalog |
| Insights            | Manual dashboard observation           | Automatic insight generation                |
| Root cause analysis | Link from RED metrics to traces/logs   | Full RCA workbench with causality graph     |
| Configuration       | Plugin-specific settings               | Shared knowledge graph configuration        |

## How it works

Application Observability powered by the knowledge graph works in three stages:

1. **Instrument**: instrument your applications with OpenTelemetry SDKs or Grafana Beyla and send telemetry data to Grafana Cloud.
2. **Activate**: enable the Applications dataset and build the knowledge graph to discover your services and generate insights.
3. **Observe**: use the services catalog, service detail pages, and RCA tools to monitor and troubleshoot your applications.

## Pricing

Application Observability uses host-hours pricing. Hosts are identified automatically from OpenTelemetry resource attributes. For details, refer to:

- [Host-hours pricing](/docs/grafana-cloud/observe-and-act/monitor-applications/application-observability-kg/pricing/): how host hours are counted and how hosts are identified.
- [Optimize costs](/docs/grafana-cloud/observe-and-act/monitor-applications/application-observability-kg/optimize-costs/): techniques to reduce costs without losing visibility.
- [knowledge graph pricing](/docs/grafana-cloud/knowledge-graph/get-started/knowledge-graph-pricing/): additional costs associated with the knowledge graph platform.