# Clusters

Clustering in Mixpeek is a warehouse-native grouping operation: the multimodal equivalent of SQL GROUP BY, allowing you to group similar documents together based on feature similarity rather than exact field matches.

[Cluster your own content](https://studio.mixpeek.com/namespaces/create?mode=managed&utm_source=homepage&utm_medium=clusters_hero&utm_campaign=mvs_onramp)[Read Documentation](https://mixpeek.com/docs/enrichment/clusters)

### Key Concepts

Vector-Based Clustering

\- Semantic similarity grouping

\- Embedding-based clustering

\- Advanced algorithms (HDBSCAN, K-Means)

Attribute-Based Grouping

\- Metadata-based organization

\- Time-based grouping

\- Custom field clustering

## Overview

Clustering enables you to organize and group documents based on their feature similarity. Unlike traditional SQL GROUP BY operations that group rows based on exact field matches, clustering uses similarity metrics to group documents that share similar characteristics.

### When to Use Clusters vs. Taxonomies vs. Ontologies

**Use Clusters when:**

You need to discover and group similar content automatically. Perfect for finding patterns, duplicate detection, content organization, and recommendations based on similarity. Works without predefined categories or relationships.

**Use Taxonomies when:**

You need to classify content into predefined, known categories. Best when you have established classification systems (e.g., product categories, content types).

**Use Ontologies when:**

You need to understand relationships between entities and traverse those connections. Ideal for knowledge graphs and multi-hop reasoning.

💡 **Together, they enrich retrieval:** Taxonomies classify your content, Ontologies model relationships between entities, and Clusters group similar items-all working to make your multimodal data more searchable, organized, and intelligent.

![t-SNE visualization of multimodal document clusters showing Sports Content, Product Catalog, News & Media, Training Docs, and Medical Imaging collections](/assets/mixpeek-clusters-tsne.svg)

## Clustering Types

Mixpeek supports various clustering approaches through the Grouper interface.

### Vector Clustering

Groups documents based on embedding similarity using algorithms like K-means or DBSCAN

*   •Perfect for finding visually or semantically similar content
*   •Supports multiple clustering algorithms
*   •Configurable similarity thresholds

### Categorical Clustering

Groups documents based on detected categories, objects, or topics

*   •Organize content by subject matter
*   •Group by detected objects or entities
*   •Support for hierarchical categories

## What You Can Achieve

Real outcomes from implementing clustering in your multimodal data pipeline.

### Automatic Discovery

Find similar content instantly without manual tagging. Discover duplicate products, related videos, or near-identical documents in seconds.

### Smart Recommendations

Boost engagement by 2-3x with similarity-based suggestions. Show users "more like this" without knowing anything else about the content.

### Hidden Patterns

Uncover trends and anomalies automatically. Identify emerging topics, unusual outliers, or content categories you didn't know existed.

Ready to learn how clustering works under the hood?

[Read the Technical Documentation](/docs/enrichment/clusters)

## Ready to Get Started?

Start organizing and grouping your multimodal content with Mixpeek clustering today.

[Schedule Demo](/demo)[Read Documentation](/docs/enrichment/clusters)