The robots meta directive max-image-preview: large asks Google to show an image preview sized up to the full viewport width when the page appears in search results. A meta robots generator that supports current Google tokens can emit this setting as part of a validated robots meta tag for HTML pages and as an equivalent X-Robots-Tag response header for non-HTML resources such as PDFs, images, or video files. The value large is one of only three documented settings for that directive, alongside none and standard, and it is the most permissive of the three. Because the image directive sits next to text-focused directives like max-snippet and nosnippet, a generator that refuses contradictory combinations is useful when you only want to widen the image preview without changing the text snippet behavior. The output is exact syntax; whether Google actually serves the larger preview still depends on crawl access, any structured data permissions already on the page, and the engine's own processing schedule.
What max-image-preview: large actually controls
The max-image-preview directive is a separate channel from text snippet controls. Google documents three values, and large is the one that requests the widest preview. None asks for no image preview to appear at all, standard asks for a default-sized preview thumbnail, and large asks for an image preview that can be sized up to the full viewport width of the searcher's device. The directive only governs how Google may render image previews in its search results page for the URL; it does not change which images get indexed, where the image is stored, or how the source image file itself is delivered to a browser.
The setting is independent of nosnippet and max-snippet, which control text and video preview behavior. You can combine max-image-preview: large with nosnippet to hide the text excerpt while still asking for a wide image thumbnail, and Google's documentation states that nosnippet does not by itself prevent a static image thumbnail from appearing. The two preview families are evaluated separately, so a generator that refuses to combine nosnippet with a numeric max-snippet will still happily combine nosnippet with max-image-preview: large.
How the three image preview values compare
| Value | What it requests | Typical use |
|---|---|---|
| none | No image preview shown in the result | Pages where image previews would leak confidential context |
| standard | A default-sized preview thumbnail | Generic articles where a basic preview is acceptable |
| large | Preview sized up to viewport width | Photo-led articles, product galleries, hero-image news pieces |
The difference between standard and large shows up on queries where the user is on a wide device. A standard preview sits at a constrained width regardless of screen; a large preview scales toward the viewport edge, which can matter when the dominant image carries most of the story, like a recipe photo, a product shot, or a hero illustration. None is the choice when the image on the page is decorative, low-resolution, or contains text you do not want surfaced outside your own site.
How to build a robots tag with max-image-preview: large
- Open the Meta Robots Generator and pick the crawler target that matches your policy. The general robots option emits a meta name="robots" tag; googlebot or googlebot-news restricts the rule to Google's text or news surface.
- Tick only the directives that match your resource policy. For an image-led article where you want the widest preview, turn on max-image-preview and select large from the value picker.
- Add any supporting directives you actually want, such as max-snippet: -1 to allow Google to choose a flexible text length, or noimageindex if you do not want the on-page images added to Google Images. Leave everything else unchecked so the output stays minimal.
- Copy the generated HTML meta element into the head of the page, or copy the X-Robots-Tag header text into your server, application, or CDN configuration. The generator produces both outputs from the same choices.
- Deploy with crawl access still open, then fetch the public URL as the intended crawler would. Inspect the served HTML and the response headers to confirm only one copy of the directive is present and no other system has added a conflicting rule.
Choosing HTML meta vs X-Robots-Tag for the large preview
The HTML meta element is the right channel when you control the markup of an HTML page and can place the tag inside the head. The X-Robots-Tag HTTP header is the right channel when the resource is not HTML, such as a PDF, image file, or video, or when your deployment pipeline already manages response headers more reliably than page templates. Both forms deliver the same directive set; the difference is where the tag travels.
Pasting the header text into the page body has no effect. Headers must be set by the web server, application, or CDN. If you use the meta robots generator for a PDF download, apply the X-Robots-Tag on the actual response that delivers the PDF, not on the HTML page that links to it. For a single image embedded in an HTML article, the page-level meta tag covers the image's preview behavior within that result; the image file itself only needs a header directive if you serve it standalone, for example as the canonical version referenced in an image sitemap.
For a cheat-sheet view of every directive the generator accepts, the Meta Robots Generator cheat sheet on directives and syntax lays out the full token list alongside its accepted values.
Where large can be superseded by other permissions
The max-image-preview directive is not a universal switch. Google's robots meta tag documentation states that these preview limits can be overridden when a publisher has separately granted specific permission through structured data, AMP relationships, or licensing arrangements. If your page uses Article or Product schema with image properties you have explicitly licensed for reuse, Google may serve a wider preview than the directive requests.
The same hierarchy applies to the text directives. nosnippet still trumps any numeric max-snippet value, which is why the generator refuses to emit both at once. For image previews, the practical implication is that adding max-image-preview: large does not guarantee a large thumbnail on every query; it sets the default ceiling that Google will respect absent other permissions. If a guarantee matters for a specific page, that permission has to come from the structured data side or the AMP relationship side, not from the meta tag side alone.
Verifying the deployed tag and watching for conflicts
After deployment, search engines process changes on their own crawl schedule, so verification is part of the workflow. Fetch the public URL with a tool that shows response headers, then inspect both the HTML source and the headers. A CMS, SEO plugin, reverse proxy, or CDN can each inject their own robots meta tag or X-Robots-Tag header, so you may see more than one declaration in the response.
When Google encounters conflicting rules, it applies the more restrictive one. A page with both max-image-preview: large and an older tag that says max-image-preview: standard would land on standard. Inventory the delivered output before pushing another tag; remove duplicates at the source rather than adding another directive to outvote them. Once the single canonical declaration is in place, wait for the next crawl and re-check the result.
A related point worth noting: a crawler has to be allowed to fetch the resource before it can read either form of the directive. If robots.txt disallows crawling, Google may never see the max-image-preview: large rule, so the URL can remain known from external links while the page content is not processed at all. To make the new preview effective, the resource must be fetchable, and then the directive has to reach the crawler on that fetch.
If you're weighing options, Robots.txt Generator Alternative: Skip the Server covers this in detail.