Is llms.txt required for Google AI Overviews? No
No. Google does not document llms.txt as a requirement for appearing in Google AI Overviews or AI Mode. For a Dubai, UAE, UK or Liverpool service business, the sensible starting point is much less exotic: can Google crawl the page, index it, show a snippet for it, and find genuinely useful content on it?
That answer is less exciting than buying an AI optimisation add-on, admittedly. It is also where the useful work tends to be. A missing optional file is an easy thing to point at in an audit. A service page that is blocked, thin, duplicated, noindexed or vague about what the business actually does takes more effort to sort out.
Google’s documented position is narrower than the sales pitch
Google’s guidance on AI features in Search says there are no additional technical requirements for AI Overviews or AI Mode beyond the usual requirements for indexed content that is eligible to appear with a snippet in Google Search. Meeting those requirements does not guarantee that a page will appear in an AI result. It means the page is eligible to be considered.
That distinction matters. Eligibility is not inclusion, and inclusion is not a promise of traffic, enquiries or a particular answer format. Search results vary by query, available sources and Google’s assessment of what helps the user.
It also applies specifically to Google Search AI features. Other AI products may publish their own access methods, crawler controls or content conventions. Do not take a Google-specific answer and stretch it across every platform because somebody has put AI in the heading.
What llms.txt does not prove
llms.txt is a proposed convention intended to provide structured guidance for language models. Some tools and providers may choose to use it. Its presence does not prove that Google can index your key service page, that the page is snippet-eligible, or that Google will use it in an AI Overview.
Nor does its absence, by itself, explain why an SME is missing from Google AI results. That is a rather tidy diagnosis for a problem with several moving parts.
An illustrative audit might flag a missing llms.txt file as the reason a consultancy is absent from AI answers. Meanwhile, its main services page redirects through an old campaign URL, the live version has a noindex instruction left from staging, and the copy says little beyond broad claims about quality. The file is not the first item on that list.
Run the ordinary Search checks before adding anything extra
Before assessing an optional convention, check whether the important commercial pages are in a usable state for ordinary Google Search. Focus on pages that explain a paid service, show relevant experience, answer common buyer questions or provide a sensible route to enquiry.
- Confirm crawl access. Googlebot needs to be able to fetch the page and its key resources. A robots.txt rule, firewall challenge or broken server response can interfere before content quality is even considered.
- Check indexing intent. Inspect the live page response and HTML for noindex instructions, incorrect canonicals and redirects that send the page somewhere unexpected. A WordPress staging setting copied into production is a boring but common cause of trouble.
- Check snippet eligibility. Google AI features rely on content that can be shown as a Search snippet. Restrictions such as nosnippet can affect this. Do not assume that being indexed alone settles the question.
- Read the page as a buyer would. Does it plainly say who the service is for, what is included, where the business operates, why it is credible and what a visitor should do next? A page made of slogans and stock phrases gives systems very little to work with.
- Check the public version on mobile. Make sure the content is actually present outside an editor preview. It is worth checking after a plugin update, cache change or page-builder deployment, especially if accordions, tabs or mobile-only sections hold important service details.
For broader context on what should be assessed across access, structure and content, see the practical scope of an AI-readiness audit. The work overlaps with SEO, but it is not simply a new name for a generic SEO checklist. This explanation of an AI-readiness audit versus an SEO audit sets out where the questions differ.
Keep crawler permissions separate from AI Overview eligibility
Another source of muddle is treating every Google crawler name as though it controls the same thing. Googlebot is relevant to Google Search crawling. Google-Extended is a separate control related to certain Gemini training and grounding uses, not a switch for AI Overview inclusion.
If your WordPress site has robots.txt rules, check what they actually permit before making broad changes. Our guide to Googlebot and Google-Extended controls in WordPress covers that narrower decision. It is adjacent to AI Overview eligibility, but it is not the same question.
The useful decision on llms.txt
Once the fundamentals are sound, you can decide whether llms.txt is worth testing as an optional addition for the platforms you care about. That decision should be based on documented support and a clear reason for maintaining it, not a vague fear that every new file is mandatory.
For many service businesses, improving a few core pages will be the better use of the budget: clearer headings, specific service descriptions, real location and sector context, accessible page structure, working internal links and proof that the enquiry route behaves properly. A contact form success message is not proof of delivery either, so check the mailbox and spam handling while you are there.
Can a page qualify and still miss an AI Overview?
Yes. A page can be crawlable, indexed and eligible for snippets without appearing in a particular AI Overview. Google does not guarantee inclusion, and results can change by query, user context and the information available. The practical aim is to remove avoidable technical barriers and publish material that answers the underlying question properly.
Check the platform requirement before buying the proposed fix. If you want a grounded review rather than another file added for the sake of it, Ask for an AI-readiness review tied to documented access and content requirements.