In short. noarchive restricts content use according to each service's documented rules, while Google Search treats it as a retired directive.

Definition: noarchive is an instruction in a robots meta tag, an HTML element that supplies page-level crawler settings. Its documented effects depend on the service and include model training restrictions and, historically, suppression of cached-page links.

Key facts

As of 17 October 2026, Amazon's bot documentation defines noarchive as an instruction against using the page for model training.
Bing's 22 September 2023 publisher announcement assigns noarchive restrictions on Bing Chat answers, links and model training, while preserving ordinary search eligibility.
Google's documentation changelog records the retirement of its noarchive rule on 2 October 2024, following the removal of cached links.

What does Amazon document?

Amazon's About AmazonBot page covers Amazonbot, Amzn-SearchBot and Amzn-User. It says these agents respect the page-level noarchive robots meta tag when accessing web pages. The stated restriction concerns using the page for model training.

The same documentation separately defines noindex as an instruction against indexing the page. This distinction matters when selecting a setting: the documented purpose of noarchive is a restriction on subsequent content use, whereas noindex concerns inclusion in an index.

What does Bing's announcement specify?

The Bing announcement also defines nocache, which permits URL, title and snippet display in answers. When both directives appear, Bing applies nocache. Their application across current Copilot products remains unverified here.

What remains of Google's support?

Google's robots meta tag specification lists noarchive among historical and unused rules. Google Search ignores it because the cached-link feature has ended.

The retirement notice says removal of an existing tag is unnecessary because other search engines and services may use it. An audit of an existing tag therefore needs to identify its intended recipient before deciding whether to remove it.

How is the directive written?

The directive is written in the tag's content attribute. Google's placement guidance places it in the page's <head> section. The following syntax matches the generic noarchive tag in Bing's announcement:

<meta name="robots" content="noarchive">

The name attribute identifies the recipient. The value robots addresses crawlers generally; support for the instruction still depends on each service. Inspecting the HTML confirms which instruction a page supplies, while the operator's documentation establishes its meaning.

FAQ

Does noarchive remove my page from Amazon's index?

Amazon's documentation assigns index exclusion to noindex. It assigns the model-training restriction to noarchive.

Should I delete noarchive because Google retired it?

Google's retirement notice says you can leave the tag in place because other services may use it. Review the services your tag is intended to address before changing it.

Sources

Check how a page reads to answer engines. Open the GEO score analyzer