---
title: Recommender API
description: Build custom recommendation experiences with Luigi's Box recommender endpoints.
slug: recommendations/api
docKind: hub
hub: recommendations
---

Use the Recommender API when you want to control the markup, placement logic, or delivery flow yourself. Start with the main recommender endpoint, then expand into offline publishing and customization APIs as needed.

## Endpoints

<div class="lb-guide-grid">
  <a class="lb-guide-card" href="/recommendations/api/v1/recommender/">
    <h3>Recommender API</h3>
    <p>Fetch recommendation results directly for custom widgets and backend-driven experiences.</p>
  </a>
  <a class="lb-guide-card" href="/recommendations/train-data/">
    <h3>Train Data Import</h3>
    <p>Import training data to improve recommender quality and cold-start performance.</p>
  </a>
  <a class="lb-guide-card" href="/recommendations/batch-publisher/">
    <h3>Batch Publisher</h3>
    <p>Generate recommendations offline for use cases such as newsletters and batch exports.</p>
  </a>
  <a class="lb-guide-card" href="/recommendations/customization/">
    <h3>Customizations API</h3>
    <p>Pin, block, and adjust recommendation behavior with explicit customization rules.</p>
  </a>
</div>

## Quickstarts

<div class="lb-guide-grid">
  <a class="lb-guide-card" href="/quickstart/recommendations/building-custom/">
    <h3>Build custom recommendations</h3>
    <p>Render your own recommendation UI and wire the required analytics events.</p>
  </a>
</div>

## Guides

<div class="lb-guide-grid">
  <a class="lb-guide-card" href="/quickstart/recommendations/choosing-models/">
    <h3>Choose models</h3>
    <p>Start by matching business goals and page context to the right recommendation models.</p>
  </a>
  <a class="lb-guide-card" href="/recommendations/concepts/">
    <h3>Basic concepts</h3>
    <p>Understand batching, personalization, and the request structure behind the API.</p>
  </a>
  <a class="lb-guide-card" href="/recommendations/models/">
    <h3>Reference models</h3>
    <p>Review the standard models and how they behave in different recommendation placements.</p>
  </a>
</div>
