> ## Documentation Index
> Fetch the complete documentation index at: https://docs.breadbowl.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# BreadBowl Embed (Alpha)

> Add your text and get ranked references back

BreadBowl Embed is an alpha retrieval API. Send it documents, search with natural language, and get the most relevant document and chunk IDs back.

<Note>
  Your application keeps the original text. BreadBowl Embed returns ranked references, scores, and metadata—not embedding vectors or generated answers.
</Note>

<CardGroup cols={2}>
  <Card title="Try it in five minutes" icon="rocket" href="/quickstart">
    Create an index, add one document, and run your first search.
  </Card>

  <Card title="Browse the API" icon="square-terminal" href="/api-reference/overview">
    See every endpoint, request field, and response.
  </Card>
</CardGroup>

## How it works

1. **Create an index** for a collection of documents.
2. **Add documents** using IDs from your own system.
3. **Search the index** and use the returned IDs to load the original content.

If you only need to rank a small set of documents once, use [transient scoring](/guides/transient-scoring) instead of creating an index.

## Before you start

The production API is available at `https://embed.breadbowl.ai`. You also need the API key supplied during onboarding. Keep the key in a backend or trusted development environment—never in browser or mobile code.

During the alpha, use BreadBowl Embed only with non-regulated data. See [data and API key safety](/security/data-handling) before using real content.