# Understanding Amazon’s Website Crawlers

> Canonical source: [Understanding Amazon’s Website Crawlers](https://www.answer.cloud/pages/did-you-know-amazon-runs-its-own-ai-bots-hosted/)

Use Amazon’s published crawler purposes rather than inferring model training or product use from a request.

Amazon publishes information about its website crawlers and how site owners can recognize them and express crawl preferences. Consult the current [Amazonbot documentation](https://developer.amazon.com/amazonbot) for agent purposes and verification.

## Avoid inferring more than the evidence shows

An Amazonbot request is not sufficient evidence that a particular page trained a model or powered a specific Amazon product. Search, user-triggered retrieval, and other crawling purposes should be distinguished using the documented agent and available request data.

The Amazon Q console screenshot below illustrates a product interface. It does not identify a crawler or prove a connection between a logged request and that product.

## Measure your own website

Review eligible requests, status codes, and classification limits. Verify identity where the provider supplies a supported method. Keep crawler activity separate from human referrals and answer citations.

See [agent analytics categories](https://www.answer.cloud/pages/what-bots-does-answer-cloud-track-hosted/) and [visibility metrics](https://www.answer.cloud/pages/llm-visibility-metrics-the-kpi-playbook-for-seo-and-marketing-leaders-hosted/).


![Article illustration](https://www.answer.cloud/assets/cms/article-23612-260c3f0b7c7d.png)