# Embeddings

The embeddings API generates vector representations of text. Use embeddings for semantic search, document similarity, clustering, and Retrieval-Augmented Generation (RAG).

## What Are Embeddings

Embeddings convert text into numerical vectors that capture semantic meaning. Similar texts produce similar vectors, enabling:

* **Semantic search** - Find content by meaning, beyond keyword matching
* **Document similarity** - Compare how related two pieces of content are
* **Clustering** - Group similar content together
* **RAG** - Retrieve relevant context for AI responses

## IAIEmbeddingService

The primary interface for embedding operations:

{% code title="IAIEmbeddingService.cs" %}

```csharp
public interface IAIEmbeddingService
{
    // Single embedding
    Task<Embedding<float>> GenerateEmbeddingAsync(
        string value,
        EmbeddingGenerationOptions? options = null,
        CancellationToken cancellationToken = default);

    Task<Embedding<float>> GenerateEmbeddingAsync(
        Guid profileId,
        string value,
        EmbeddingGenerationOptions? options = null,
        CancellationToken cancellationToken = default);

    // Multiple embeddings
    Task<GeneratedEmbeddings<Embedding<float>>> GenerateEmbeddingsAsync(
        IEnumerable<string> values,
        EmbeddingGenerationOptions? options = null,
        CancellationToken cancellationToken = default);

    Task<GeneratedEmbeddings<Embedding<float>>> GenerateEmbeddingsAsync(
        Guid profileId,
        IEnumerable<string> values,
        EmbeddingGenerationOptions? options = null,
        CancellationToken cancellationToken = default);

    // Advanced: Get the underlying generator
    Task<IEmbeddingGenerator<string, Embedding<float>>> GetEmbeddingGeneratorAsync(
        Guid? profileId = null,
        CancellationToken cancellationToken = default);
}
```

{% endcode %}

## Basic Usage

{% code title="SearchService.cs" %}

```csharp
using Microsoft.Extensions.AI;
using Umbraco.AI.Core.Embeddings;

public class SearchService
{
    private readonly IAIEmbeddingService _embeddingService;

    public SearchService(IAIEmbeddingService embeddingService)
    {
        _embeddingService = embeddingService;
    }

    public async Task<float[]> GetEmbedding(string text)
    {
        var embedding = await _embeddingService.GenerateEmbeddingAsync(text);
        return embedding.Vector.ToArray();
    }
}
```

{% endcode %}

## Choosing a Profile

### Default Profile

Once you have [configured a default embedding profile](/ai-in-umbraco/backoffice/managing-settings.md) in the backoffice, you can call without specifying a profile:

```csharp
var embedding = await _embeddingService.GenerateEmbeddingAsync(text);
```

### Specific Profile

Pass the profile ID:

```csharp
var embedding = await _embeddingService.GenerateEmbeddingAsync(profileId, text);
```

## Common Use Cases

### Semantic Search

1. Generate embeddings for all searchable content
2. Store embeddings in a vector database
3. When searching, embed the query and find nearest neighbors

### Content Similarity

Compare two pieces of content:

```csharp
var embedding1 = await _embeddingService.GenerateEmbeddingAsync(content1);
var embedding2 = await _embeddingService.GenerateEmbeddingAsync(content2);

var similarity = CosineSimilarity(embedding1.Vector, embedding2.Vector);
```

### RAG (Retrieval-Augmented Generation)

1. Embed user question
2. Find relevant content via vector search
3. Include retrieved content in chat prompt

## In This Section

{% content-ref url="/pages/dEgVXLBZwfTzdVO1Cf4r" %}
[Generating Embeddings](/ai-in-umbraco/using-the-api/embeddings/generating-embeddings.md)
{% endcontent-ref %}

{% content-ref url="/pages/Udl0yL1Fj2uw1Wclze5F" %}
[Batch Embeddings](/ai-in-umbraco/using-the-api/embeddings/batch-embeddings.md)
{% endcontent-ref %}


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.umbraco.com/ai-in-umbraco/using-the-api/embeddings.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.