Web Design

AI search visibility: no llms.txt file needed

AI search visibility: no llms.txt file needed

No. You do not need an llms.txt file, or any other special AI markup, for Google's AI Overviews or AI Mode. Google's Search Central guide says Google Search does not use llms.txt files, and that keeping one will neither harm nor help visibility or rankings in Google Search [1]. There are no extra technical requirements beyond the rules that already apply to ordinary Google Search [2]. AI search visibility on Google still starts with a page that can be indexed and shown with a snippet, not with a new file at the root of the site [1][2].

What the llms.txt proposal is

The filename comes from a public proposal, not from a Google rule. Version 2, on the proposal's own site, asks for a markdown file named llms.txt so software agents can get a short guide to a site [6]. Jeremy Howard is named as the author. The page lists September 3, 2024 as the first publication date and August 10, 2026 as the modification date [6].

The file can sit at the site root or on a deeper path, with a title, a short summary, and links to markdown copies of important pages [6]. The only required section is a top heading with the site or project name [6]. The author describes it as something an agent opens while helping a person, expects use at answer time rather than as training material, and says it sits beside robots.txt and a sitemap because those files do different jobs [6]. The write-up calls itself an informal overview, open to community input [6].

It does not say Google Search reads the file, and it does not say a business must publish one to appear in Google's AI answers [6]. The page says the files are used most heavily for software documentation, and that a business could use the same shape to outline its structure [6]. That hope is for agents that choose to fetch the file, not a switch inside Google Search.

What Google requires for AI Overviews and AI Mode

Google's page on AI features and your website, last updated December 10, 2025, says there are no additional requirements to appear in AI Overviews or AI Mode, and no other special optimizations are necessary [2]. A supporting link has to be indexed, eligible for a snippet in Google Search, and in line with the Search technical requirements [2]. There are no additional technical requirements [2]. The July 10, 2026 guide adds that the site must also be included in Search generative AI features in Search Console [1]. Inclusion is the default [5]. Both documents say that meeting the rules does not mean Google will crawl, index, or serve the page [1][2].

The July 2026 guide says these features are rooted in Google's core ranking and quality systems [1]. Google describes a grounding step that retrieves pages from the Search index with those systems, then reviews the pages and shows links [1]. It also describes query fan-out: related searches the model issues so it can fetch more results [1]. The December 2025 page says that pattern can show a wider set of links than one classic search, and that AI Overviews often do not trigger, because they appear only when Google's systems decide they add something [2]. Allowing Googlebot to crawl, keeping important content in text, and linking the page from the rest of the site are ordinary habits on that page, not a special AI setup [2].

What Google says about the file

The July 2026 guide lists llms.txt files and other special markup among things you can ignore for Google Search [1]. You do not need new machine-readable files, AI text files, markup, or markdown to appear in Google Search, including its generative AI features, because Google Search does not use them [1]. Google may crawl many kinds of files besides HTML, and that does not mean the file gets special treatment [1]. Keeping an llms.txt file for some other service will neither harm nor help visibility or rankings in Google Search, because Google Search ignores it [1]. The same page says to skip chopping pages into tiny pieces for AI, and to skip unnecessary AI text files such as llms.txt [1]. No third-party tool has access to Google's internal ranking or AI systems, so the guide says to be wary of tools that promise ranking success on that basis [1].

The December 2025 page says you do not need new machine-readable files, AI text files, or markup for these features, and that there is no special schema.org structured data to add [2]. The July 2026 guide says structured data is not required for generative AI search [1]. If you already use it, that guide says it can still help with eligibility for ordinary rich results, and the AI features page says the markup should match the visible text [1][2]. That is not a file to add for AI Overviews, and it is not a list of types that will get a page mentioned. Names for special work aimed at answer engines and at generative search are common online, and Google says many of those shortcuts are not supported by how Search works [1].

How to read this in Search Console

Two reports cover this. Google's AI features documentation, last updated December 10, 2025, says appearances in AI features are included in overall search traffic in Search Console, inside the Performance report's Web search type [2]. The July 2026 guide tells owners to use the Generative AI performance report [1]. That help page says that, as of August 31, 2026, the insights are rolled out to all websites worldwide, and that the report includes data from the Web search type [3]. It is a breakout of impressions, not a sign that those appearances left the main report [2][3].

The dedicated report shows impressions for AI Overviews and AI Mode [3]. An impression is a time a link to your site was shown in a generative AI feature on Google Search [3]. You can group by page, country, device, or date, and page rows use the final URL after redirects, mostly the canonical URL [3]. A filter separates text searches from image searches, including Google Lens, Circle to Search, uploaded images, and Chrome image search [3].

On the chart, two links from the same site in one feature count as one impression, while a page grouping can count them separately, so the totals can disagree [3]. A missing report usually means too few impressions, or that the property was excluded [3]. Search Labs experiments are left out, and the table stops at 1,000 rows [3]. AI Overviews often do not trigger, so a quiet report is not proof that a file is missing [2].

Clicks and position are on the methodology page the AI features documentation links [2][4]. A click in an AI Overview or in AI Mode is a click that leaves Google for an outside page [4]. An AI Overview occupies one position, shared by every link inside it, so that number is not a rank among those links [4]. A follow-up question in AI Mode is a new query [4]. Query fan-out is a separate reason a page can be retrieved for a neighboring question [1]. Standard impression rules generally count an item already on the current page of results, while a block that scrolls or expands on its own typically counts only once the item is in view or expanded, and Google says those habits can change [4].

Check the generative AI report by page for impressions [3], then the same URLs in the Web Performance report for clicks [2]. A gap can be the impression threshold or an excluded property, not a missing file [3].

The control that changes inclusion

Participation is a Search Console setting under Settings, then Search generative AI, and as of August 31, 2026 that control is rolled out to all websites worldwide [5]. It covers AI Overviews, AI Mode, and generative AI features in Google Discover, which has its own report, while the Search report above is the one for AI Overviews and AI Mode [3][5].

Include is the default. Content can appear as links and as material that helps ground responses, and the help page says the site can receive impressions and traffic from those features [5]. That is eligibility, not a forecast. Exclude keeps your links and content out of those features, including grounding, and you then will not receive traffic or impressions from them [5]. Other sites can still appear, and their content may look similar to yours [5]. A property can inherit the setting from a parent [5].

The control is not a ranking or inclusion signal for the rest of Search, and it does not override Merchant Center or Google Ads [5]. noindex is the documented way, on that same page, to stay out of Google Search entirely [5]. The toggle does not affect AI training. The help page points to Google-Extended if you want to limit training of the models used for responses in these features [5]. The AI features page, separately, points to Google-Extended for limiting training and grounding in some of Google's other systems [2].

A change generally takes effect within one to two days, though caching can make some pages linger [5]. Child properties inherit unless someone overrides them, so check which property is open if the report looks wrong [5]. Page-level limits are separate: robots.txt for Googlebot controls crawling, and nosnippet, data-nosnippet, max-snippet, or noindex limit what is shown [2].

What we do instead of adding the file

We would not add an llms.txt file in the hope of showing up in Google's AI answers. Google's guide says Google Search ignores it, and that keeping one for another product neither helps nor hurts Google Search [1].

For Google, the checks are ordinary. Pages should be indexed, Googlebot should be allowed to crawl them, important content should be in the text, and snippets should be allowed [2]. Leave the Search generative AI control on Include unless you mean to leave those features [5]. Google's guide puts unique, useful content ahead of the other suggestions on that page, and it says many near-duplicate pages made to manipulate generative AI responses violate the scaled content abuse policy [1]. An up-to-date Business Profile, and Merchant Center if you sell products, can still help those details show in AI responses and in other Google results [1].

We are a Charlotte-based team. On our Forged Sites plans, every plan includes a monthly search review covering ordinary search, answer engine optimization, and generative engine optimization. When Search Console and Analytics are connected, that review can use them. It is a set of recommendations, not a guarantee of an AI recommendation, a ranking, or a sale.

Practical takeaways

  • You do not need an llms.txt file, or other special AI markup, for Google's AI Overviews or AI Mode. Google Search does not use those files [1].
  • Keeping the file for another tool will neither help nor hurt visibility or rankings in Google Search [1].
  • Eligibility is ordinary: indexed, allowed to show a snippet, and included in Search generative AI features. Include is the default, and indexing is not guaranteed [1][2][5].
  • Use the Generative AI performance report for impressions in AI Overviews and AI Mode. Those appearances also sit in the Web Performance report [2][3].
  • An AI Overview is one position shared by its links. A follow-up in AI Mode counts as a new query [4].
  • To leave the features, use the Search generative AI control. That toggle is not noindex, and it is not the training control [5].
  • A missing report usually means too few impressions, or that the property was excluded, not that a file is missing [2][3].

How we can help

Have more questions or want to get in touch? Visit https://ideaforgestudios.com/contact-us-idea-forge-studios/ , call (980) 322-4500, or email [email protected].

Citations

  1. Google Search Central, "Optimizing your website for generative AI features on Google Search" (2026-07-10)
  2. Google Search Central, "AI features and your website" (2025-12-10)
  3. Google Search Console Help, "Generative AI performance report (Search)" (2026-08-31)
  4. Google Search Console Help, "What are impressions, position, and clicks?" (2026)
  5. Google Search Console Help, "Search generative AI control" (2026-08-31)
  6. llms-txt, "The /llms.txt file, v2" (2026-08-10)
Tags: AI Overviews Google Search ai mode llms.txt search console
Our Strongest Offering

Forge Your Next Website

Forged Sites are custom-built, static-first websites with a full AI content engine on board — no CMS to log into, no plugins to break, no builder to fight.

  • Working target: WCAG 2.2 AA
  • During work hours, an account manager still reviews material changes
  • DraftDash auto-drafted blogs keep your content engine running
  • Ethel AI-powered forms filter spam and capture genuine leads