What page data does the Site API return?

It can return raw Open Graph and Twitter Card fields, values inferred from HTML, and a merged `hybridGraph` result.

Why might I not see a request for every preview?

OpenGraph.io allows cached results and documents a default maximum cache age of five days. The calling customer can choose different cache settings.

Does OpenGraph.io Bot respect robots.txt?

Compliance is not confirmed in the bot metadata, and Cloudflare's bot directory marks this identity as not following robots.txt. Treat a rule as a preference unless your own logs show otherwise.

Is OpenGraph.io Bot an AI training crawler?

No training role is documented. The API has an optional AI-assisted metadata feature, but this crawler's recorded purpose is fetching pages for API results and link previews.

Can I verify it from the `OpenGraph.io` token?

No. The entry has no verified IP status or signature directory, and OpenGraph.io can use proxies. A matching header is useful for classification, not authentication.

OpenGraph.io Bot & AI Search: What to Know

What is OpenGraph.io Bot?

OpenGraph.io Bot appears when an OpenGraph.io customer asks the service to inspect a URL, often so a consumer product can display a shared-link preview. Its Site API reads Open Graph fields, Twitter Card data, and ordinary HTML metadata, then returns a merged result to the calling product.

The API can serve cached data, with a documented default maximum cache age of five days. For a fresh or failed lookup it may render JavaScript, retry the request, or use a proxy according to the customer's options and the current API defaults. Those behaviors mean one preview action does not always map neatly to one origin request.

OpenGraph.io also offers an optional AI enhancement for metadata extraction. That option does not turn this bot into a general AI search crawler, and the record gives no foundation-model training purpose. A fetch can affect the preview returned to the API customer, but it does not establish eligibility for AI answers or citations.

The stable token is OpenGraph.io; Cloudflare's bot directory records a full OpenGraph.io/1.1 crawler header. Robots.txt compliance is unsettled: the bot metadata leaves it unknown, while Cloudflare's bot directory marks the identity as not following robots.txt. The record has no verified IP or signature, so neither the token nor a proxy address proves who initiated a request.

Indirectly relevant

Is OpenGraph.io Bot relevant for AI search?

Indirectly. OpenGraph.io Bot has no AI product of its own, but its output can end up in the systems that AI answers draw on.

OpenGraph.io Bot fetches pages to build link-preview cards. It does not feed an AI index, but it reads the same Open Graph and structured metadata that AI systems parse to understand your pages, so clean preview metadata doubles as AI-friendly metadata.

How to handle OpenGraph.io Bot

Allow the bot if products using OpenGraph.io should be able to build previews of your public URLs. Check that useful titles, descriptions, and images are present in the server response, and keep private pages behind authentication.

You can publish a preference for the stable token:

User-agent: OpenGraph.io
Disallow: /

Do not rely on that rule for a hard block. this bot does not confirm compliance, and Cloudflare says the bot does not follow robots.txt. Use rate limits, edge rules, or application controls when access must stop, while accounting for OpenGraph.io's documented proxy and retry options.

Examples

A messaging product calls the OpenGraph.io Site API for an article and receives a title, description, and image assembled from the page metadata.
A preview lookup returns cached metadata, so the publisher sees no new origin request until the API needs a fresher result.
A protected site returns `403`, and the API retries through another proxy while the site continues to enforce the same access policy.

Frequently asked questions about OpenGraph.io Bot

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

An OpenGraph.io customer submits a target URL to one of the service's APIs. Shared-link previews are a common use, but the API also supports curation and monitoring tools.