Why this matters to you
llms.txt has been marketed as the new robots.txt for AI — a simple file you add to your website to help AI engines understand what you offer. Hotel tech vendors have started pushing it. SEO plugins are auto-generating it. Hospitality blogs are calling it essential for 2026.
We wanted to know two things: how many hotels have actually implemented it, and whether it makes any difference to how AI engines like ChatGPT, Perplexity, and Gemini find and recommend those hotels.
The answers are: not many, and not yet.
We found zero requests from major AI crawlers to llms.txt files in server log analysis. The standard is real — the adoption isn’t.
Key findings at a glance
01
5.8% adoption across 104,214 hotels
Only 6,044 of the hotel websites we scanned have an llms.txt file. The companion llms-full.txt is even rarer at 0.4%. The standard has not reached hospitality in any meaningful way.
02
WordPress SEO plugins drive a third of it
34.2% of llms.txt files were auto-generated by plugins — AIOSEO (20.1%), Yoast SEO (10.8%), Rank Math (3.3%). Most hotels that have it didn’t consciously decide to implement it.
03
8.1% are using it wrong
One in twelve hotels with an llms.txt file is using it as access-control rules, like robots.txt. Those files provide zero value to AI engines — and may confuse crawlers that encounter unexpected directives.
04
AI engines aren’t reading it
Server log analysis across a sample of hotels shows zero requests from GPTBot, PerplexityBot, Googlebot, or ClaudeBot to llms.txt files. The standard is aspirational, not operational.
What this means for your hotel
llms.txt is not moving the needle on AI visibility today. OpenAI, Google, Anthropic, and Perplexity have not indicated that their crawlers use it as a signal for search or recommendation. Hotels implementing it are making a bet on future adoption — not gaining present advantage.
The hotels with llms.txt do score 61% higher on Schema.org quality (22.1 vs 13.7 average). But that correlation runs the other way: llms.txt adoption is a proxy for overall technical SEO maturity, not the cause of better visibility. Improving your Schema.org markup directly would have more impact than adding llms.txt on top of weak structured data.
The right order of operations
Fix Schema.org first. Get your robots.txt right for AI crawlers. Structure your content for machine reading. Then add llms.txt as cheap, low-risk future-proofing — not as the primary intervention.
What to do about it
1. Don’t prioritize llms.txt over Schema.org.
Schema.org is actively used by AI engines today. The median hotel Schema.org quality score in our dataset is 0 out of 100 — that is the gap that matters for current AI visibility. llms.txt addresses a problem that isn’t causing you harm yet.
2. If you add llms.txt, make it a proper site index.
Only 41.3% of hotels with llms.txt follow the intended spec as a structured index of their site. The file should list your key pages with brief descriptions — not marketing copy, not robots.txt directives. A well-structured llms.txt takes under an hour to write and costs nothing to host.
3. If you’re on WordPress, check what your plugin generated.
If you use AIOSEO, Yoast, or Rank Math, you may already have an llms.txt you didn’t know about. Check yourdomain.com/llms.txt. If it exists, review whether it accurately represents your property or contains generic plugin output.
The evidence
Finding 1 — Adoption by country
The US leads adoption, driven partly by higher WordPress market share and a more active tech-SEO community. France significantly trails, consistent with broader patterns of slower AI tooling adoption in French hospitality.
llms.txt adoption rate by country
104,214 hotel websites scanned · April 2026
Spain\u2019s above-average rate is partially driven by a single local hotel platform that auto-generates llms.txt for its clients.
Finding 2 — How files are generated
The majority of llms.txt files in hospitality are plugin-generated, not intentionally authored. This has direct implications for quality — generic plugin output rarely contains the property-specific detail that would make the file useful to an AI engine.
llms.txt generation source
Among 6,044 hotels with llms.txt
“Custom-built” includes files that could not be attributed to a known generator. Quality within this group varies significantly.
Finding 3 — Content quality breakdown
Among hotels that have llms.txt, most files fall short of the standard’s intended purpose. Only a small fraction contain the kind of rich, property-specific content that would actually help an AI engine.
llms.txt content quality classification — among 6,044 hotels with the file
| Classification | Share | Description |
|---|---|---|
| Proper site index | 41.3% | Follows spec: lists pages with descriptions |
| Generic plugin output | 34.2% | Auto-generated, minimal property-specific content |
| Misused as robots.txt | 8.1% | Contains Disallow/Allow directives — provides zero value |
| Rich hotel descriptions | 3.1% | Includes amenities, policies, room types, links |
| Other / unclassified | 13.3% | Empty, malformed, or non-standard format |
61%
Schema.org quality lift for hotels with llms.txt. Hotels that have llms.txt score 61% higher on Schema.org quality on average (22.1 vs 13.7). The correlation reflects overall technical SEO investment — not a causal effect of llms.txt itself.
On AI crawler activity
Server log analysis was conducted across a sample of hotels that shared access logs. We checked for requests from GPTBot, OAI-SearchBot, PerplexityBot, Googlebot, Google-Extended, ClaudeBot, CCBot to the /llms.txt path. Zero requests were recorded across the entire sample during the observation period (March–April 2026). This is consistent with public statements from OpenAI, Google, and Perplexity, none of whom have announced active llms.txt consumption.
Frequently asked questions
Yes, but only after fixing more impactful issues. If your Schema.org markup is incomplete or uses the wrong type, fix that first. llms.txt takes 30 minutes to implement correctly and costs nothing — it’s reasonable future-proofing once the higher-priority work is done.
No, and confusing them is the most common mistake. robots.txt tells crawlers which pages they may access. llms.txt is a plain-text summary of your site intended to help AI engines understand what you offer. It has no enforcement mechanism — it’s informational only.
The standard is less than two years old and has no formal backing from any major AI provider. OpenAI, Google, Anthropic, and Perplexity have not announced support for it in their crawlers or inference pipelines. Adoption may come, but it hasn’t yet.
Insurance against future adoption. The cost is low (30 minutes), and if major AI engines do start consuming it, hotels that already have a well-structured file will have an immediate advantage. The risk of not having it, once it matters, outweighs the effort of adding it now.
At minimum: your hotel name, a one-sentence description, and links to your key pages (rooms, amenities, location, policies) with brief descriptions of each. Ideally: check-in/out times, amenity list, cancellation policy summary. Avoid marketing copy — write for a machine that wants facts, not feelings.
How we ran the study
104,214
Hotels scanned
7
Countries
6,044
llms.txt files found
15s
Request timeout
We fetched /llms.txt and /llms-full.txt from 104,214 reachable hotel websites during April 2026. Hotel URLs were sourced across seven countries: United States, United Kingdom, France, Germany, Spain, Italy, and the Netherlands.
For each file found, we ran generator identification (matching known plugin signatures), content classification (site index vs. robots-style vs. rich description vs. other), and file size distribution analysis.
Schema.org quality scores were computed independently using a 15-property three-tier scoring system, then correlated against llms.txt presence. AI crawler activity was analyzed via server log access from a subset of participating hotels.
Limitations. Hotel URL sources may over-represent properties with stronger web presence. Server log analysis covers a non-random sample. Country coverage reflects where hotel URL data was available at scale.
Want to know how visible your hotel is to AI engines today?
Huxo’s AI Visibility Report audits your hotel across ChatGPT, Perplexity, and Google AI Mode — and tells you which signals are missing and what to fix first.