Noarchive

Google's current documentation lists noarchive under historical and unused rules — it's ignored on Search. Bing still uses it to control Chat/Copilot answer inclusion and AI-training opt-out. Syntax, the 2024 Google timeline, and when it still matters.

First published: Jul 2, 2026 · Last updated: Jul 19, 2026 · Advanced
demand #13 in Meta Tags#44 in On-Page#212 in Technical SEO#290 on the site

Noarchive is a robots meta directive (`<meta name="robots" content="noarchive">`) that once removed Google's 'Cached' link. Google's current documentation lists `noarchive` under 'historical and other unused rules' and says it's ignored, because the cached-link feature no longer exists (retired Feb 2024; the `cache:` operator followed by Sept 2024). Google's Search Liaison said in 2024 the rule would still be respected, but Google's own current docs say otherwise — treat it as inert on Google today, not a quietly-honored control. It's not access control either, and can't stop scraping, browser/CDN caching, screenshots, or third-party archiving. The real reason to still add it is Bing: since September 2023, `noarchive` opts your content out of Bing Chat/Copilot answers and Microsoft's AI training, while `nocache` is a lighter-touch variant. It's distinct from `noindex` (removes the page from results) and `nosnippet` (hides the preview snippet). Use the `X-Robots-Tag: noarchive` header for non-HTML files.

TL;DR — noarchive (<meta name="robots" content="noarchive">, or the X-Robots-Tag: noarchive header for non-HTML files) once removed Google’s “Cached” link. Google retired cached pages in Feb 2024 and the cache: operator by ~Sept 2024, so Google now lists noarchive under “historical and other unused rules” — it’s inert on Google’s SERPs. The differentiating reason to still care is Bing: since Sept 2023, noarchive opts content out of Bing Chat/Copilot answers and Microsoft’s AI training, while nocache is a lighter-touch variant. It’s distinct from noindex (drops the page) and nosnippet (hides the preview). It does not stop the Wayback Machine.

Evidence for this claim Google lists noarchive as a historical rule it no longer uses because cached links are no longer available in Google Search. Scope: Current Google Search behavior only. Confidence: high · Verified: Google Search Central: Historical robots rules Evidence for this claim Microsoft documents noarchive as preventing use of page content in Bing Chat/Copilot answers, links from those answers, and generative AI foundation-model training while leaving ordinary Bing web-search visibility available. Scope: Microsoft Bing/Copilot controls; distinct from Google Search, noindex, and independent web archives. Confidence: high · Verified: Bing Webmaster Blog: New controls for webmasters

What noarchive is — a serving directive, not an index control

noarchive is one of the robots meta directivesThe robots meta tag is an HTML element in a page's head — <meta name=\"robots\" content=\"noindex\"> — that tells search engines how to index and serve that page. It's crawl-then-obey: a page blocked in robots.txt is never fetched, so the tag is never seen.. Its historical job was narrow: stop search engines from keeping a cached/archived copy of the page available to searchers. It never removed the page from the indexStoring a crawled page in the search index so it can appear in results. Crawled is not the same as indexed — Google selects what to keep, and indexing isn't guaranteed., never touched rankings, and never suppressed the snippet — it only governed the cached copy.

The syntax is a standard robots meta tagThe robots meta tag is an HTML element in a page's head — <meta name=\"robots\" content=\"noindex\"> — that tells search engines how to index and serve that page. It's crawl-then-obey: a page blocked in robots.txt is never fetched, so the tag is never seen.:

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

You can scope it to a single engine’s crawlerA crawler — also called a spider or bot — is an automated program that fetches web pages, extracts their links, and queues new URLs to visit. Search engines use crawlers to discover and download content for their index. instead of all of them — <meta name="googlebot" content="noarchive"> for Google only, or <meta name="bingbot" content="noarchive"> for Bing only — and you can combine it with other directives in one tag (e.g. content="noarchive, nosnippet").

The X-Robots-Tag header — for PDFs and other non-HTML files

A <meta> tag only works where there’s an HTML <head> to put it in. For a PDF, an image, or any non-HTML resource, you send the same directive as an HTTP response header instead. Google’s spec is explicit that this works for any rule: “The X-Robots-Tag can be used as an element of the HTTP header response for a given URL. Any rule that can be used in a robots meta tag can also be specified as an X-Robots-Tag.” So the header form is:

X-Robots-Tag: noarchive

This is the same delivery pattern as noindex and the rest of the robots family — head meta tag for HTML, HTTP header for everything else.

Why noarchive does nothing on Google anymore

Google removed cached pages. The timeline is worth pinning down because almost no competing article connects it back to what it means for this specific tag:

  • February 2024 — Google removed the “Cached” link from search results.
  • ~September 2024 — Google finished killing the cache: search operator, per Search Engine Land’s follow-up coverage. Roughly a seven-to-eight-month wind-down.

Because there’s no cached feature left, noarchive has nothing to act on. Google now files it under “the following rules aren’t used by Google Search and are ignored. We’re including these rules here because people have often asked about them or we used them in the past” — its “historical and other unused rules” section. On noarchive specifically: “The noarchive rule is no longer used by Google Search to control whether a cached link is shown in search results, as the cached link feature no longer exists.”

The adjacent nocache rule gets the same treatment — “The nocache rule isn’t used by Google Search” — which matters because people constantly treat the two as synonyms (they diverge on Bing; see below).

What Sullivan said in 2024 — and what Google’s current docs say now

Here’s a nuance worth being precise about, because two Google-sourced statements don’t fully agree. When the cache link was going away, Google’s Search Liaison Danny Sullivan addressed noarchive directly: “You’re going to see cache: go away in the near future, too. But wait, I hear you ask, what about noarchive? We’ll still respect that; no need to mess with it. Plus, others beyond us use it.” That was Sullivan’s on-the-record position in September 2024.

But Google’s own current documentation is more direct than that quote: “The following rules aren’t used by Google Search and are ignored.” “Ignored” is a stronger claim than “we’ll still respect that,” and the two statements don’t fully square up. The governing source is Google’s own current documentation, not a two-year-old quote — so I’m not going to tell you noarchive is quietly honored on Google. Treat it as a rule Google’s Search docs say it doesn’t act on today.

Practically, the outcome barely changes either way: whether Google “ignores” the rule outright or once “respected” it with zero visible effect, there’s nothing on Google’s SERPs noarchive controls anymore, and no reason to add it there for Google’s sake. The difference matters for how you describe the tag accurately — not for what you should do about it.

noarchive vs noindex vs nosnippet vs max-snippet

These get muddled constantly. They live in the same robots family but do completely different jobs:

DirectivePage still indexed?Snippet still shown?What it controls
noarchiveYesYesThe cached/archived copy (inert on Google; live on Bing for AI)
noindexNoN/ARemoves the page from search results entirely
nosnippetYesNoSuppresses the text snippet / video preview
max-snippet:0YesNoCaps snippet length to zero — functionally like nosnippet

Google’s own definitions: noindex is “Do not show this page, media, or resource in search results”, and nosnippet is “Do not show a text snippet or video preview in the search results for this page” — and notably, that same nosnippet rule now also “prevent[s] the content from being used as a direct input for AI OverviewsAI Overviews are the AI-generated summary box Google shows above or within its regular search results, written by Gemini models from pages retrieved out of Google's normal Search index. It's a Search feature, not a separate platform or index. and AI Mode.” If your goal is “don’t show any preview text on Google,” nosnippet or max-snippet:0 (Google: “0: No snippet is to be shown. Equivalent to nosnippetnosnippet is a robots directive (meta tag or X-Robots-Tag header) that suppresses all snippet text and the video preview for a whole page; data-nosnippet is an HTML attribute on span, div, or section that excludes just that section from the snippet.) is the tool — not noarchive, which never touched snippets in the first place.

Where noarchive still earns its keep: Bing and AI

This is the real 2026 story, and it’s the angle almost nobody covers. Bing gives noarchive a current, meaningful function that has nothing to do with the old cache link.

Per Bing’s September 2023 announcement about controlling how content is used in Bing Chat, the tags now break down roughly like this:

  • noarchive-tagged content is not included in Bing Chat/Copilot answers, isn’t linked to there, and isn’t used to train Microsoft’s generative AI foundation models.
  • nocache-tagged content may still appear in Bing Chat answers, but only as URL / title / snippet — and those elements may still be used in training.
  • If a page carries both tags, Bing treats it as nocache (the less restrictive of the two — genuinely counterintuitive, so worth noting).
  • Content with neither tag is fair game for both Bing Chat answers and AI training.
  • Either way, normal Bing web-search visibility is untouched — the page still ranks in Bing’s regular organic results.
Evidence for this claim Microsoft documents noarchive as preventing use of page content in Bing Chat/Copilot answers, links from those answers, and generative AI foundation-model training while leaving ordinary Bing web-search visibility available. Scope: Microsoft Bing/Copilot controls; distinct from Google Search, noindex, and independent web archives. Confidence: high · Verified: Bing Webmaster Blog: New controls for webmasters Bing’s live help page for its robots tags renders client-side and resisted automated verification, so the specific noarchive/nocache behavior above is described from Bing’s own September 2023 blog post rather than quoted verbatim; confirm the exact wording in a browser before treating any single sentence as a direct Bing quote. That announcement is also the only primary source behind this section — it documents Bing’s policy as of September 2023, not an independently reverified 2026 confirmation across every current Copilot, partner, or captioning surface. Nothing foundA 302 (\"Found\") is a temporary redirect: it forwards users to a new URL while telling search engines the original URL should stay in the index. It's a weak canonicalization signal, not the zero-equity dead end of SEO folklore. in this research contradicts it, but treat it as “documented policy since 2023” rather than a guarantee that behavior is identical everywhere Bing content appears today.

So on Bing this is the cruxChrome User Experience Report — Google's public dataset of real-world (field) performance data from eligible Chrome users. It's the official field-data source behind the Core Web Vitals program., reframed: the question noarchive answers is no longer just “may users view a cached copy?” — it’s “may Bing reuse this content inside AI chat answers and use it to train models?” That’s the differentiator, and it’s why I lead with Bing rather than the dead Google cache link.

When to still use noarchive in 2026

  • To opt out of Bing’s AI/chat reuse and training — the single strongest, most current reason. If you don’t want your content feeding Bing Chat/Copilot answers or Microsoft’s model training, noarchive is the lever.
  • Time-sensitive pages, as a convention-following courtesy — promos that expire or pages you’d rather crawlersA crawler — also called a spider or bot — is an automated program that fetches web pages, extracts their links, and queues new URLs to visit. Search engines use crawlers to discover and download content for their index. that still honor noarchive (or nocache) not keep an archived copy of. This is not a security control: it does nothing to stop the live page from being viewed, screenshotted, copied, cached by a browser or CDN, or archived by tools that don’t honor the directive. For paywalled or genuinely sensitive content, use real access control (authentication, IP restriction) — not noarchive.
  • As harmless legacy — if it’s already in your templates, leaving it costs you nothing on Google and does a real job on Bing.

For most sites, noarchive is boilerplate carried over from a pre-2024 mental model — and that’s fine. It becomes genuinely useful again only when you actively want to control AI reuse on Bing, or as a defensive habit on content you’d rather not see archived by tools that still respect the tag.

Myths worth clearing up

  • noarchiveNoarchive is a robots meta directive that tells search engines not to keep a cached copy of a page. Google's current documentation lists it under historical and unused rules and says it's ignored, because the cached-link feature it once controlled no longer exists — but on Bing it still controls whether content is used in Bing Chat/Copilot answers and AI training. It is not an access-control measure. still hides a Cached link on Google.” No — the cached feature is gone for everyone, tag or not. Google’s docs say the rule “is no longer used…as the cached link feature no longer exists.”
  • “noarchive is a dead tag, delete it.” Not urgent either way — Google’s docs now say the rule is ignored, but leaving it costs you nothing there, and it has live effect on Bing’s AI surfaces. Removing it can matter on Bing even though it won’t matter on Google.
  • “noarchive protects sensitive or private content.” No — it’s a request to search engines, honored inconsistently, and does nothing to prevent the origin page from being viewed, screenshotted, copied, cached by a browser/CDN, or archived by a tool that doesn’t honor it. Use actual access control for anything you need to keep private.
  • “noarchive affects rankings.” There’s no documented ranking effect — it’s a serving-behavior rule, not a ranking signal. You’ll see a widely repeated claim attributed to John Mueller that it doesn’t impact rankings; I couldn’t trace that to a verifiable original source, so I won’t present it as a quote — but the directional claim (no ranking impact) is consistent with how Google documents the tag.
  • “noarchive stops the Wayback Machine.” No — it’s a search-engine directive, not a universal do-not-archive instruction. The Internet Archive and other third-party archivers don’t act on it.
  • “noarchive and nocache are the same everywhere.” Historically yes, and still true as a baseline synonym for classic cachingCaching stores a copy of a page or resource — in a browser, a CDN edge node, or a search crawler's own cache — so it can be served again without regenerating or re-downloading it. It isn't a direct ranking factor, but it feeds page speed and crawl efficiency. — but no longer true on Bing, where they now carry different AI-reuse meanings.

Where noarchive fits with everything else

noarchive sits in the robots meta family alongside noindex (its index-control cousin) and nosnippet / max-snippet (the snippet controls), all delivered through the robots meta tag in the head or the X-Robots-Tag header for non-HTML files — the same delivery mechanics covered across the meta tagsMeta tags are HTML elements in a page's head that pass metadata about the page to search engines and browsers. For SEO only a few matter — the title element, the meta description, and the robots meta tag — while meta keywords and most others are ignored. hub. The one-line summary: on Google it’s a documented, ignored historical rule with no visible effect; on Bing it’s a live, current control over AI-answer inclusion and model training; on neither engine is it a security or access control. Keep those realities separate and noarchive stops being confusing.

Add an expert note

Pin an expert quote

New person? Create their unclaimed profile at /admin/experts/ → Pin a quote first.