llms.txt Explained
llms.txt is a proposed Markdown file that a website places at its root, such as example.com/llms.txt, to give AI tools a short summary of the site and a curated list of its most useful pages. It is a community proposal, not an adopted standard, and major search engines may not use it at all.
- What it is: A plain text file written in Markdown, with a title, a summary and lists of links to key pages.
- Why it was proposed: Web pages are full of menus, scripts and ads, so a clean list of the important pages is easier for a language model to use.
- What it is not: It is not robots.txt, so it cannot allow or block crawlers, and it is not a sitemap for search engines either.
- Status: Google has said it does not use llms.txt for Search, and other AI companies have not clearly confirmed using it.
- Effort: Writing one takes an hour or two for a small site, so it is a cheap experiment rather than a priority.
This tutorial follows one example: a JEE and NEET coaching institute in Kota with a website that has course pages, a fee structure page, a scholarship test page, admission steps and a blog of exam preparation tips. The owner has heard that llms.txt helps with AI search and wants to know what it is, whether it is worth doing, and how to create one correctly.
Prerequisites for Creating an llms.txt File
- Access to the site's root folder: You need to upload a file so it appears at the main domain, through the hosting panel, FTP, the site builder's file manager or a plugin.
- A list of your key pages: Know which 10 to 30 pages best explain what the business offers, such as courses, fees, admissions and contact.
- Basic Markdown: A hash (#) makes a heading, a greater-than sign (>) makes a quote, and a link is written as text.
- The basics done first: Check that the site is crawlable and indexed, as covered in technical SEO. llms.txt does nothing for pages that crawlers cannot reach.
Setup: Understand the llms.txt Format
The proposal describes a simple structure, in this order:
- An H1 heading: The name of the site or organisation, which is the only required part of the file.
- A blockquote: A short summary of the site with the key information someone needs to understand the rest of the file.
- Optional notes: Plain paragraphs or lists with extra details, such as who the site is for.
- H2 sections: Each one holds a list of links, written as "- Page name: a short note on what the page contains".
- An "Optional" section: Links that are useful but can be skipped when an AI tool has limited space.
The proposal also suggests two related ideas: a longer llms-full.txt file that contains the full text of key pages, and clean Markdown versions of pages at the same address with ".md" added. Both are optional, and most small businesses can skip them.
Step-by-Step: Create and Publish llms.txt
Step 1: Decide Whether It Is Worth Doing
Be honest about the expected benefit: llms.txt is most common on developer documentation sites, where programmers may point AI coding tools at it directly, while for a local business or a coaching institute there is no proven effect on AI answers. It is still low effort, so the institute decides to treat it as a small experiment after the more important work in how to get cited by ChatGPT is done.
Step 2: Choose the Pages
List the pages that answer the questions students and parents ask most. For the institute, these are the JEE and NEET course pages, the fee structure, the scholarship test, admission steps, batch timings, hostel information, results pages that show only verified results, and contact details. Blog posts that are not central go in the Optional section, while login pages, the student portal and anything private are left out entirely.
Step 3: Write the Summary
In two or three sentences, state what the institute is, where it is, who it serves and what makes it different, using facts only. Avoid slogans like "No. 1 coaching institute", because they add nothing a model can use and may count as misleading advertising claims.
Step 4: Write the File
Here is a complete example for the institute. Replace every address and detail with your own real pages and facts.
# Example Academy, Kota
> Example Academy is a coaching institute in Kota, Rajasthan, that prepares Class 11, Class 12 and dropper students for JEE Main, JEE Advanced and NEET UG. It offers classroom batches in Kota and live online batches, with Hindi and English medium options.
Key facts:
- Courses: JEE (two year, one year and dropper batches) and NEET UG (two year, one year and dropper batches)
- Admission: through the Example Academy Scholarship Test or direct enrolment
- Contact: admissions desk, Monday to Saturday, 9 am to 6 pm
## Courses
- [JEE Main and Advanced two year programme](https://www.example.com/courses/jee-two-year): syllabus, batch timings, study material and fee summary
- [NEET UG one year programme](https://www.example.com/courses/neet-one-year): syllabus, weekly test schedule and doubt sessions
- [Online live batches](https://www.example.com/courses/online): class timings, recordings and device requirements
## Admissions and Fees
- [Fee structure](https://www.example.com/fees): fees for every course, instalment options and refund policy
- [Scholarship test](https://www.example.com/scholarship-test): eligibility, syllabus, dates and how scholarships are calculated
- [Admission steps](https://www.example.com/admissions): documents needed and how to enrol
## Student Life
- [Hostel and PG guidance](https://www.example.com/hostel): approved hostels, safety rules and parent contact policy
- [Student support](https://www.example.com/support): counselling and mentoring for students living away from home
## Optional
- [Exam preparation blog](https://www.example.com/blog): study plans and subject tips
- [About the institute](https://www.example.com/about): history and faculty profilesStep 5: Upload It to the Root
Save the file as llms.txt in plain UTF-8 text and upload it so it opens at the main domain, for example www.example.com/llms.txt, not inside a subfolder. On WordPress, some SEO plugins can create and update the file for you. On Shopify and other hosted builders, check whether the platform lets you place files at the root.
Step 6: Test It
Open the address in a browser and confirm that the file loads as plain text, not as a styled page or an error, and that every link in it opens the correct page. Also make sure robots.txt does not accidentally block the file itself.
Step 7: Keep It in Sync and Watch the Logs
Whenever a course, fee page or address changes, update llms.txt in the same week, because an outdated file points AI tools to wrong information. Ask the developer or hosting provider to check the server logs every month for requests to /llms.txt, and note which user agents fetch it. Those log entries are the only direct evidence of whether any AI tool actually reads the file.
Common Mistakes
- Expecting guaranteed results: llms.txt is a proposal, and there is no evidence that it guarantees citations in any AI assistant.
- Using it to block crawlers: It cannot block anything, because access rules belong in robots.txt, as explained in technical SEO.
- Listing every page: A file with hundreds of links defeats the purpose of a short, curated guide.
- Marketing language: "Best results in India" gives a model nothing useful and may be a misleading claim.
- Private links: Never include student portals, internal documents or pages with personal data.
- Letting it go stale: Old fees or closed batches in the file can end up in AI answers.
- Neglecting the basics: Clear pages, crawler access and genuine mentions matter much more, as explained in generative engine optimization.
Next Steps
- Structured data: Describe the institute and its courses to search engines with schema markup.
- Measure: Track whether AI answers mention the institute before and after changes, using the method in measuring AI search visibility.
- Technical side: Learn how AI search tools retrieve pages in how AI search engines work.
How AI Changes Site Guidance Files Like llms.txt
What AI Automates Now
AI assistants can read a sitemap or a list of URLs, suggest which pages belong in llms.txt, and draft the one line notes for each link. Some site plugins and developer tools now generate the file automatically from the site's structure.
What Still Needs a Human
A person must decide which pages truly represent the business, confirm that every fact in the summary is current, and remove anything private, because only the institute knows which fee page is final and which batch has closed.
Risk to Watch
Generated files can include draft pages, test URLs or pages with personal data, and AI-written summaries can add claims the institute cannot prove. Treat the file as a public document, because anyone on the internet can open it.
Do It with AI
Use this prompt to draft an llms.txt file from your list of key pages. It works in ChatGPT, Claude or Gemini.
You are helping a business in India create an llms.txt file, following the proposed format: one H1 with the site name, a short blockquote summary, optional key facts, H2 sections with Markdown link lists, and an Optional section. Business: [name, city, what it offers, who it serves] Key pages: [paste 10 to 30 URLs with the title of each page and one line on what it contains] Less important pages: [paste URLs that can go in the Optional section] 1. Write the H1 and a factual summary of two or three sentences. 2. Group the key pages under clear H2 headings. 3. Write one factual note for each link, based only on what I told you. 4. List anything I must confirm before publishing. Do not add pages I did not give you, and do not invent results, rankings, fees or claims.
- Export or list the site's key pages with their titles.
- Run the prompt and check every link and note against the live page.
- Upload the file to the site root and open it in a browser to test it.
- Put a monthly reminder in the calendar to update it and check the server logs.
Check Before You Use It
- Facts: Every fee, date and course detail must match the live website and the institute's official documents.
- Brand fit: Keep the summary factual and calm, without slogans or rankings.
- Compliance: Include no private or student data, and avoid result or ranking claims that the institute cannot prove, in line with the rules on misleading coaching advertisements.
Quick Quiz
Pick an answer to check yourself. Nothing is saved.
Question 1 / 3
1. The Kota coaching institute wants to stop an AI crawler from fetching its student portal. Which file should it change?
Frequently Asked Questions
What is llms.txt?
llms.txt is a proposed Markdown file placed at the root of a website, such as example.com/llms.txt. It gives AI tools a short summary of the site and a curated list of its most useful pages, written in a format that language models read easily.
Is llms.txt an official standard?
No. It is a community proposal, not a standard adopted by a standards body or by the major search engines. Some tools and developer documentation sites use it, but there is no guarantee that any particular AI assistant reads it.
Does Google use llms.txt?
Google has said publicly that it does not use llms.txt for Search, including its AI features. Google's AI Overviews rely on its normal crawling and indexing, so robots.txt, sitemaps and good pages matter far more.
What is the difference between llms.txt and robots.txt?
robots.txt tells crawlers which parts of a site they may or may not fetch, and major crawlers follow it. llms.txt does not allow or block anything; it only suggests which pages are most useful and summarises the site.
Can llms.txt hurt my SEO?
A simple, accurate llms.txt file is unlikely to cause harm, since search engines treat it as an ordinary text file. The risks are spending time on it instead of work that matters more, and letting it go out of date so it points to wrong pages.
Related Articles
- How to Get Cited by ChatGPT, Perplexity and GeminiHow to get cited by ChatGPT, Perplexity and Gemini: a seven step framework, a page checklist and a monthly routine, shown with a GST billing app example.
- What is GEO (Generative Engine Optimization)Generative engine optimization for marketers: how AI assistants pick the brands they name and how to become one of them, with a D2C millet snack example.
- Technical SEOTechnical SEO explained simply: crawlability, indexing, canonicals, speed, mobile and site structure, with a Surat saree store example and an AI prompt.
- Measuring AI Search VisibilityMeasure AI search visibility with a prompt list, four simple metrics, a GA4 channel for AI referrals and a tracking sheet, with a Lucknow chikankari store.
- Schema Markup (Structured Data)Schema markup tutorial: choose a schema type, write valid JSON-LD, test it and track rich results, with a LocalBusiness example for an Indore kirana store.
- What is AEO (Answer Engine Optimization)Answer engine optimization explained: write pages that win featured snippets, People also ask boxes and voice replies, with a Hyderabad clinic example.