llms.txt for Malaysian Businesses: The Honest Guide

llms.txt takes 15 minutes to add to your website. It costs nothing. And right now, the data shows it does almost nothing for your AI search visibility. So why are 5.61% of the world’s top 10,000 websites already using it — and why should you add it anyway? That is the honest answer this guide gives you.

What Is llms.txt and Where Did It Come From?

On 3 September 2024, Jeremy Howard — co-founder of Answer.AI and fast.ai — published a proposal for a new web standard. The idea was simple: place a Markdown file at https://yourdomain.com/llms.txt that tells AI systems what your site is about and which pages matter most.

Howard’s original audience was developers. Specifically, he wanted AI coding agents like Cursor and Claude Code to navigate documentation sites more effectively. He was not designing an SEO tool. He was not designing a way to game AI chatbot citations. That distinction matters enormously when you are deciding how much time to invest in it.

The format is deliberately minimal. Your llms.txt file must contain an H1 (your site name), a blockquote (a short description of your business), and optional H2 sections listing links to your key pages with one-line descriptions. You can read the full details at the official llms.txt specification.

Adoption has grown fast. HTTP Archive data and research by Casey Burridge (published June 2026) shows 5.61% of the top 10,000 websites now have a valid llms.txt file — up from just 1.04% in July 2025. That is 5.4x growth in 12 months. Shopify silently rolled it out to every store in May 2026, which accounts for 78.1% of all Shopify sites now carrying the file. There are also 66 WordPress plugins that generate it automatically, with the “Website LLMs.txt” plugin alone logging over 40,000 active installs.

How llms.txt Differs From robots.txt and sitemap.xml

These three files serve completely different purposes, and confusing them leads to misplaced expectations.

robots.txt is about access control. It tells crawlers what they are and are not allowed to fetch. Think of it as the bouncer at the door: “You can come in here, but not there.” It directly affects indexing decisions.

llms.txt is about comprehension. It tells AI systems: “Here is what matters on this site and how to understand it.” There is no blocking, no restricting — just a curated summary. The two files coexist without conflict and serve entirely separate functions.

sitemap.xml is a full index of every page you want search engines to know about. llms.txt, by contrast, is a highlight reel. You choose the handful of pages most important for an AI to understand your business, and you describe them in plain language.

In practice, adding llms.txt does not touch your existing robots.txt or sitemap setup. You are simply adding a new file to your root directory — one that is considerably simpler to write than either of the other two.

Which AI Systems Actually Read llms.txt? (The Data, 2026)

Here is where the honest picture gets uncomfortable — and why you should read this section before spending more than 15 minutes on the task.

A 90-day audit by Limy.AI analysed over 500 million AI bot visits across a sample of websites. Across that entire dataset, only 408 requests fetched /llms.txt. That is 0.00008% of total AI bot activity. Every major crawler — GPTBot, ClaudeBot, PerplexityBot, Google-Extended — overwhelmingly skips the file entirely and reads your HTML directly.

The citation data is equally flat. SE Ranking studied 300,000 domains and found zero statistical correlation between having llms.txt and AI citation frequency. A separate Trakkr study of 37,894 cited domains found that sites with llms.txt averaged 6.8 AI citations versus 6.7 for sites without — a difference so small it carries a p-value of 0.85, which is statistically indistinguishable from noise.

Furthermore, growth has slowed. Casey Burridge’s June 2026 research noted that month-on-month adoption flattened in May and June 2026. The initial wave of early adopters has passed.

Where llms.txt does demonstrably work is in its original use case: AI coding agents. Tools like Cursor, Claude Code, and Windsurf, when pointed at a documentation site with a well-structured llms.txt, navigate it more efficiently. If your business runs a developer-facing product or documentation hub, the file earns its keep. For most Malaysian service businesses, however, that scenario does not apply.

How to Add llms.txt to Your Malaysian Website in 15 Minutes

Despite the limitations above, Casey Burridge puts it well: “A minimal llms.txt is 15 minutes of work. The downside of not having one is close to zero right now. The potential upside, as AI agents and retrieval systems mature, is a cleaner signal for how your content gets understood and surfaced.”

That is the correct framing. Do it once, do it properly, then focus your energy on the things that actually drive AI visibility — which the final section of this guide covers.

Option 1 — WordPress Plugin (Easiest)

If your site runs on WordPress, search for “Website LLMs.txt” in your plugin directory. Install it, activate it, and the plugin generates a valid llms.txt file automatically at your root domain. It pulls your pages and posts into the correct format without any manual editing. With over 40,000 active installs, it is the most battle-tested option available. Review the output once it is live by visiting https://yourdomain.com/llms.txt in your browser, then edit any descriptions that sound too generic.

Option 2 — Create and Upload Manually

Open any plain-text editor. Write your content in Markdown format, following the structure shown in the next section. Save the file as llms.txt with no subdirectory path. Upload it to your web server’s root directory — the same folder that contains your robots.txt file. For most cPanel or Cloudflare-managed hosts in Malaysia, this means the public_html folder. Verify it is live by visiting the URL directly in your browser.

For non-WordPress platforms such as Webflow or custom-built sites, most platforms now allow static file uploads to the root. The upload itself takes under five minutes — the writing is where you invest the time.

A Sample llms.txt for a Malaysian Service Business

Below is a template you can adapt directly. Replace the bracketed placeholders with your actual content. Keep each description to one concise sentence.

# Your Business Name

> [Your Business Name] is a [type of business] based in [city], Malaysia,
> serving [target customers]. We specialise in [key services].

## Services

- [Service 1](https://yourdomain.com/service-1/): Brief one-line description
- [Service 2](https://yourdomain.com/service-2/): Brief one-line description
- [Service 3](https://yourdomain.com/service-3/): Brief one-line description

## About

- [About Us](https://yourdomain.com/about/): Background, team, and credentials
- [Contact](https://yourdomain.com/contact/): Enquiry form and office details

## Optional

- [Blog](https://yourdomain.com/blog/): Industry insights and guides

A few practical notes for Malaysian businesses. First, write your blockquote description in plain, factual English. Mention the city and state — Kuala Lumpur, Penang, Johor Bahru — because geographic specificity helps AI systems build an accurate entity understanding of your business. Second, link to your highest-value service pages, not your homepage. Third, if you publish content in Bahasa Malaysia as well as English, note that in your description. Multilingual signals matter more than most businesses realise — the data below explains why.

What llms.txt Won’t Do — Setting Honest Expectations

This is the section most guides skip. Read it before you tell your client or your team that you have “optimised for AI search.”

Google Search ignores it — confirmed

In the June 2026 Search Central update, Google stated directly: “Doing so won’t harm (nor help) your visibility or rankings in Google Search, as Google Search ignores them.” John Mueller went further, comparing llms.txt to the long-defunct meta keywords tag — a file that exists, that some sites maintain, but that search engines have no obligation to act on.

There is one nuance worth knowing. Google Chrome’s Lighthouse tool added “Agentic Browsing” audits in 2026, and one of those audits checks for llms.txt. However, that check is about machine-readability for AI agents interacting with your page — not about Google Search ranking. Two different Google teams are answering two different questions. Do not conflate them.

No proven effect on AI chat citations yet

As the Trakkr and SE Ranking studies confirm, having llms.txt does not meaningfully increase how often AI systems like ChatGPT, Claude, or Perplexity cite your business. The bots largely ignore the file and parse your HTML directly. Consequently, if your goal is to appear more frequently in AI-generated answers, llms.txt is not the lever to pull right now.

That said, this could change. As AI retrieval systems mature and agent-based browsing becomes more common, a well-structured llms.txt may carry more weight. The file is a long-term bet, not a short-term tactic. Add it, leave it, and revisit it in 12 months.

What actually moves the needle for AI visibility in Malaysia

If AI search visibility is a genuine priority for your business in 2026, these are the areas where evidence shows real impact — specifically in Malaysia and Singapore.

Entity schema markup is the single largest gap in the Malaysian market. Businesses with properly deployed entity schema receive 4.8 times more local AI citations than those without it. Yet under 4% of Malaysian SMEs have entity schema in place. If you want to understand where to start, our guide on schema markup for AEO covers the implementation priorities for service businesses.

The Malaysia-Singapore citation gap is real and widening. Current data puts Singapore’s AI citation rate at 58.7% versus Malaysia’s 42.1%. The gap is not about technology access — it is driven by content quality, schema deployment, and entity clarity in published content. Our GEO playbook for AI search citations breaks down exactly what is behind this gap and what Malaysian businesses can do to close it.

Bahasa Malaysia content produces a 67.3% higher AI citation rate for businesses publishing in both English and BM versus English alone. If your customers search in Bahasa Malaysia and your site is English-only, you are leaving a significant share of AI visibility on the table. This is one of the most underutilised advantages available to Malaysian businesses right now.

E-E-A-T signals remain the foundational layer. An updated Google Business Profile, author credentials on published content, and structured data that clearly identifies your business as a local entity all contribute to how AI systems build confidence in citing a source.

For a fuller picture of how Malaysian businesses can track and improve their AI search presence, see our practical guide on how to track AI search visibility in Malaysia and our analysis of why businesses don’t appear in Google AI answers. The AEO Malaysia 30-day action plan covers the full implementation sequence in one place.

Add your llms.txt today. It genuinely takes 15 minutes, and the downside of skipping it is low but nonzero. However, do not stop there. The businesses that will be cited first in AI-generated answers a year from now are the ones investing in schema, multilingual content, and entity clarity — not just the ones who uploaded a Markdown file and called it done. For a structured approach to AI search readiness, the RANK GENERAL protocol covers everything the data shows actually drives results in 2026.