> For the complete documentation index, see [llms.txt](https://docs.umbraco.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.umbraco.com/ai-in-umbraco/17.latest/management-api/embeddings/generate.md).

# Generate

Generates vector embeddings for one or more text values.

## Request

```http
POST /umbraco/ai/management/api/v1/embeddings/generate
```

### Headers

| Header          | Value                  |
| --------------- | ---------------------- |
| `Authorization` | Bearer token or cookie |
| `Content-Type`  | `application/json`     |

### Request Body

{% code title="Request" %}

```json
{
    "profileIdOrAlias": "document-embeddings",
    "values": ["Umbraco is an open-source CMS", "Content management made easy"]
}
```

{% endcode %}

### Request Properties

| Property           | Type        | Required | Description                                   |
| ------------------ | ----------- | -------- | --------------------------------------------- |
| `profileIdOrAlias` | string/guid | No       | Profile ID or alias (uses default if omitted) |
| `values`           | string\[]   | Yes      | Text values to embed (min 1)                  |

## Response

### Success (200 OK)

Returns embeddings for each input value.

{% code title="Response" %}

```json
{
  "embeddings": [
    {
      "index": 0,
      "vector": [0.0023, -0.0092, 0.0156, 0.0089, ...]
    },
    {
      "index": 1,
      "vector": [0.0089, -0.0034, 0.0201, 0.0145, ...]
    }
  ]
}
```

{% endcode %}

### Response Properties

| Property              | Type     | Description                        |
| --------------------- | -------- | ---------------------------------- |
| `embeddings`          | array    | List of embedding results          |
| `embeddings[].index`  | int      | Index of corresponding input value |
| `embeddings[].vector` | float\[] | Embedding vector                   |

### Bad Request (400)

Returned when the request is invalid or embedding generation fails.

{% code title="Response" %}

```json
{
    "type": "https://tools.ietf.org/html/rfc7231#section-6.5.1",
    "title": "Embedding generation failed",
    "detail": "No default embedding profile configured",
    "status": 400
}
```

{% endcode %}

### Not Found (404)

Returned when the specified profile doesn't exist.

{% code title="Response" %}

```json
{
    "type": "https://tools.ietf.org/html/rfc7231#section-6.5.4",
    "title": "Not Found",
    "status": 404
}
```

{% endcode %}

## Example

### cURL

```bash
curl -X POST "https://your-site.com/umbraco/ai/management/api/v1/embeddings/generate" \
  -H "Authorization: Bearer {token}" \
  -H "Content-Type: application/json" \
  -d '{
    "profileIdOrAlias": "document-embeddings",
    "values": ["Umbraco CMS", "Content management"]
  }'
```

### C# HttpClient

{% code title="Example" %}

```csharp
using var client = new HttpClient();
client.DefaultRequestHeaders.Authorization =
    new AuthenticationHeaderValue("Bearer", token);

var request = new
{
    profileIdOrAlias = "document-embeddings",
    values = new[] { "Umbraco CMS", "Content management" }
};

var response = await client.PostAsJsonAsync(
    "https://your-site.com/umbraco/ai/management/api/v1/embeddings/generate",
    request);

var result = await response.Content.ReadFromJsonAsync<EmbeddingResponseModel>();

foreach (var embedding in result.Embeddings)
{
    Console.WriteLine($"Input {embedding.Index}: {embedding.Vector.Length} dimensions");
}
```

{% endcode %}

### JavaScript

{% code title="Example" %}

```javascript
const response = await fetch("/umbraco/ai/management/api/v1/embeddings/generate", {
    method: "POST",
    headers: {
        Authorization: `Bearer ${token}`,
        "Content-Type": "application/json",
    },
    body: JSON.stringify({
        profileIdOrAlias: "document-embeddings",
        values: ["Umbraco CMS", "Content management"],
    }),
});

const result = await response.json();
result.embeddings.forEach((e) => {
    console.log(`Input ${e.index}: ${e.vector.length} dimensions`);
});
```

{% endcode %}

## Use Cases

### Semantic Search Index

Generate embeddings for content to enable semantic search:

{% code title="Example" %}

```csharp
public async Task IndexContentAsync(IContent content)
{
    var text = $"{content.Name} {content.GetValue<string>("bodyText")}";

    var response = await _httpClient.PostAsJsonAsync(
        "/umbraco/ai/management/api/v1/embeddings/generate",
        new { values = new[] { text } });

    var result = await response.Content.ReadFromJsonAsync<EmbeddingResponseModel>();
    var embedding = result.Embeddings[0].Vector;

    // Store in vector database
    await _vectorStore.UpsertAsync(content.Key, embedding, new
    {
        contentId = content.Id,
        name = content.Name,
        url = content.GetUrl()
    });
}
```

{% endcode %}

### Content Similarity

Find similar content by comparing embeddings:

{% code title="Example" %}

```csharp
public async Task<IEnumerable<IContent>> FindSimilarContentAsync(
    string text,
    int count = 5)
{
    // Generate embedding for search text
    var response = await _httpClient.PostAsJsonAsync(
        "/umbraco/ai/management/api/v1/embeddings/generate",
        new { values = new[] { text } });

    var result = await response.Content.ReadFromJsonAsync<EmbeddingResponseModel>();
    var queryVector = result.Embeddings[0].Vector;

    // Query vector database for similar content
    var similar = await _vectorStore.SearchAsync(queryVector, count);

    // Load Umbraco content
    return similar
        .Select(s => _contentService.GetById(s.ContentId))
        .Where(c => c != null);
}
```

{% endcode %}

### Batch Processing

Process content in batches for efficiency:

{% code title="Example" %}

```csharp
public async Task IndexAllContentAsync(IEnumerable<IContent> contents)
{
    const int batchSize = 100;

    foreach (var batch in contents.Chunk(batchSize))
    {
        var texts = batch.Select(c =>
            $"{c.Name} {c.GetValue<string>("bodyText")}").ToList();

        var response = await _httpClient.PostAsJsonAsync(
            "/umbraco/ai/management/api/v1/embeddings/generate",
            new { values = texts });

        var result = await response.Content.ReadFromJsonAsync<EmbeddingResponseModel>();

        for (int i = 0; i < result.Embeddings.Count; i++)
        {
            var content = batch.ElementAt(i);
            var vector = result.Embeddings[i].Vector;
            await _vectorStore.UpsertAsync(content.Key, vector);
        }
    }
}
```

{% endcode %}

## Notes

* Embedding order matches input order (use `index` to correlate)
* Vector dimensions depend on the model configured in the profile
* Processing multiple values in one request is more efficient than individual calls
* Store embeddings in a vector database for production semantic search


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## 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, and the optional `goal` query parameter:

```
GET https://docs.umbraco.com/ai-in-umbraco/17.latest/management-api/embeddings/generate.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

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.