Does Pinterest Bot respect robots.txt?

Yes. Pinterest documents support for allow, disallow, and crawl-delay rules under `User-agent: Pinterestbot`. Crawl delays above 1 second are treated as 1 second.

Does Pinterest use this crawl data to train Pinterest Canvas?

No. Pinterest states that content crawled by Pinterestbot is not used to train Pinterest Canvas, its image generation foundation model.

Why does Pinterest Bot return to a page that already has a Pin?

Pinterestbot revisits pages behind existing Pins to update details and identify broken links. That maintenance traffic can occur without a new Pin being created.

How can I verify a Pinterestbot request?

Run reverse DNS on the source address, require a hostname ending in `pinterest.com` or `pinterestcrawler.com`, and resolve that name forward to the same address. Pinterest recommends this because its addresses can change.

What happens if I block Pinterest Bot?

Pinterest will not be able to crawl the blocked pages for discovery or refresh their Pin details. Blocking this bot is a Pinterest distribution choice, not a documented opt-out from general AI training or other search engines.

Pinterest Bot: AI Crawler Guide & robots.txt

What is Pinterest Bot?

Pinterest Bot, documented by Pinterest as Pinterestbot, gathers public web content for Pinterest. Its index lets people discover material from outside websites and sends a click on a Pin back to the associated page.

The crawler handles product and article pages differently. On a product page it may collect the image, title, description, price, and availability so a Pin reflects the current offer. For an article, Pinterest displays a preview rather than reproducing the full piece. This makes the quality of page metadata relevant to what people see on Pinterest.

Pinterestbot also visits pages behind Pins that people have already created. Those requests refresh product or article details and help Pinterest find broken destination links. A repeat fetch can therefore be routine maintenance of an existing Pin, not a fresh attempt to index the whole site.

Pinterest says content fetched by this bot is not used to train Pinterest Canvas, its image generation foundation model. The operator does not describe Pinterestbot as a crawler for generative answers or general model training. Allowing it affects Pinterest discovery, Pin recommendations, and the accuracy of Pin details. It should not be presented as a route into unrelated AI search products.

Pinterest publishes several user-agent forms, including Pinterest/0.2 and Pinterestbot/1.0, all pointing to its bot page. The operator says requests come from a Pinterest network and are rate limited. US crawling uses its published 54.236.1.* range, while localized crawling can use changing addresses outside that range, so Pinterest advises against a fixed IP allowlist.

A user-agent string alone can be copied. Pinterest's crawler documentation tells site owners to verify a claimed bot with reverse DNS, check for a hostname ending in pinterest.com or pinterestcrawler.com, and then confirm that hostname resolves back to the original address. The supplied Cloudflare record has no signed-agent key directory.

Relevant for AI search

Is Pinterest Bot relevant for AI search?

Yes. Pinterest Bot collects pages for an AI product, so what it can crawl influences how AI systems describe your brand.

Pinterest Bot feeds a search index that also powers that engine's AI answer features and overviews. One crawl can serve a classic results page and a generated summary, so blocking it costs you both the rankings you would expect and a growing share of AI answers built on the same index.

How to handle Pinterest Bot

Keep Pinterestbot allowed when you want products and articles to be discoverable on Pinterest and want existing Pins to stay current. Block it when that distribution is not useful, or restrict only paths that should never supply Pin metadata.

Pinterest says it respects robots.txt. Its documented control token is Pinterestbot:

User-agent: Pinterestbot
Disallow: /

The same record can disallow a directory instead of the whole site. Pinterest also accepts Crawl-delay, but caps the effective value at 1 second. It only supports the robots file on the main domain, not a separate file on a subdomain. Cloudflare's false robots flag conflicts with Pinterest's explicit operator policy, so use the operator instructions and confirm the observed result.

Do not put confidential material behind robots.txt. Require authentication for it. If a request claims the Pinterestbot identity while ignoring a valid rule, use Pinterest's forward and reverse DNS procedure before deciding that the traffic came from Pinterest.

Examples

A retailer changes a product's price and availability, then sees Pinterestbot revisit the page to refresh the details shown on existing Pins.
A magazine permits Pinterestbot on published articles but disallows its subscriber account area, which has no useful Pin preview content.
A security analyst checks forward and reverse DNS after a high-volume client copies the Pinterestbot user agent but comes from an unrelated network.

Frequently asked questions about Pinterest Bot

Learn about AI visibility monitoring and how Promptwatch helps your brand succeed in AI search.

Pinterest says it collects page information used for Pins. Product fields can include an image, title, description, price, and availability. For articles, Pinterest uses page information to create a preview that links to the source.