# What is GraphSAGE? Definition and meaning

Learns to sample and aggregate neighbour features, so it can embed nodes it never saw in training.

## Definition

**GraphSAGE**: Learns to sample and aggregate neighbour features, so it can embed nodes it never saw in training.

In short: Inductive learning on large graphs.

## When it is used

Large graphs that change over time, such as transaction or social networks.

## Learn more

The full guide to **[GraphSAGE](/tools/graphsage)** explains how it works, when to use it and how it compares with the alternatives, with examples. The [glossary](/glossary) lists every term on the site from A to Z.

Previous term: [GraphRAG](/glossary/graphrag) · Next term: [Grok 4.7](/glossary/grok-4-7)
