llms.txt: what it is, an example, and how to create one.
llms.txt is a site-passport file for neural networks. It sits at the root (site.com/llms.txt) and briefly explains to ChatGPT, Perplexity, and Gemini what the business does: services, prices, GEO, key facts. Below — why it's needed, what's inside, a real example, and how to create your own in 5 steps.
What is llms.txt
llms.txt is a simple text file in Markdown format placed at the site's root at yoursite.com/llms.txt. Its job is to explain the essence of the business to neural networks briefly and in machine-readable form: what you do, for whom, how much it costs, where you operate, and where to look next. The idea is similar to robots.txt, but instead of access rules, this is a concise summary of the site's content, convenient for large language models.
It's important to be honest: llms.txt is not yet an official standard that every engine is required to support. It's a hygiene layer — it helps the model understand the site faster, but on its own it doesn't guarantee a top spot in answers.
Why a business needs llms.txt
A regular homepage is design, banners, scripts. Neural networks have to "fight through" the markup to understand the substance. llms.txt gives the model a clean extract of facts without the noise: name, services, prices, GEO, case studies. This reduces the risk of AI confusing what you do or citing outdated data. Combined with schema markup and content, llms.txt makes a site predictably clear to ChatGPT, Perplexity, and Gemini.
Example of llms.txt (structure)
Here's a simplified structure using our agency as an example — this is what llms.txt looks like at the root of insync.com.ua:
| Block | What it contains |
|---|---|
| # Name | H1 heading — brand name (InSync Agency) |
| > Description | One paragraph: what the business does, for whom, GEO |
| ## Key Facts | Services, prices ($450–700/mo), minimum budget, case results, GEO |
| ## Key Pages | Links to the homepage, services, niches, case studies, contacts |
| ## Contacts | Telegram, email, request form |
The main rule is facts, not marketing: specific numbers, services, and links the model can quote verbatim. This is exactly how ChatGPT, in a clean session, cites our positioning with the source "InSync Agency".
How to create llms.txt in 5 steps
Step 1. Write down the key facts: name, services, prices, GEO, 3–5 case results with numbers.
Step 2. Format it in Markdown: an H1 with the name, a description paragraph after ">", and "## Key Facts" and "## Key Pages" sections with lists and links.
Step 3. Save it as llms.txt and put it at the site's root so it opens at yoursite.com/llms.txt (text/plain type, no authorization).
Step 4. Make sure AI crawlers are allowed in robots.txt (GPTBot, ClaudeBot, PerplexityBot, OAI-SearchBot, etc.) — otherwise no one will read the file.
Step 5. Update llms.txt whenever services or prices change, and check whether neural networks accurately relay the business — this is part of the monthly GEO measurement.
Common mistakes in llms.txt
Fluff instead of facts — a model can't quote "we're the best on the market," but it can quote "$450–700/mo, 150+ projects." Outdated data — old prices or handles (check regularly). The file is blocked — behind authorization or in a subfolder the crawler can't reach. llms.txt exists, but crawlers are blocked in robots.txt — the file is useless. Relying on llms.txt alone — without content and citations, it won't pull visibility on its own.
Frequently asked questions about llms.txt
What is llms.txt in plain terms?
llms.txt is a text file at the root of a site (at site.com/llms.txt) — a kind of passport for neural networks. In machine-readable Markdown format, it briefly lays out what the business does: services, prices, GEO, key facts, and links to important pages. AI crawlers read it to quickly and accurately understand the site.
Does llms.txt guarantee appearing in ChatGPT's answers?
No. llms.txt is a hygiene, baseline layer that makes it easier for neural networks to understand a site, but on its own it doesn't guarantee mentions. It's not yet an official standard. The main role is played by content built around niche queries and external citations. llms.txt is necessary insurance, not a silver bullet.
Where should llms.txt be placed?
The file is placed in the site's root directory so it opens at https://yoursite.com/llms.txt — just like robots.txt. Optionally, you can add llms-full.txt with expanded content. Make sure the file is served with the text/plain type and is accessible without authorization.
How is llms.txt different from robots.txt and sitemap.xml?
robots.txt controls crawler access, sitemap.xml lists all URLs for indexing, and llms.txt explains the business's content in concise, machine-readable language — so the neural network understands the substance, not just the structure. These three files complement each other and are needed together.
We'll check your llms.txt and 7 other factors — for free
Our scanner takes 60 seconds to show whether you have an llms.txt, whether AI crawlers are allowed in, and how many points out of 100 your site scores for AI visibility.
See the live GEO case study