How to Create an llms.txt File: Step-by-Step for Shopify, WordPress, and Webflow

Every week, more of your audience asks ChatGPT, Perplexity, or Gemini a question your website could answer-and gets pointed somewhere else. The reason isn't bad content. It's that AI models don't read your site the way Google does. Large language models face a critical limitation: context windows are too small to handle most websites in their entirety, and converting complex HTML pages with navigation, ads, and JavaScript into LLM-friendly plain text is both difficult and imprecise.

That's the problem llms.txt was designed to solve. Proposed by Jeremy Howard in September 2024, llms.txt aims to make more efficient use of web content by artificial intelligence systems. It's a Markdown file placed at your site's root directory that gives AI a curated, structured map of your most important content. It is not a Google ranking signal, but it is increasingly used by the AI retrieval systems that now send high-converting referral traffic.

Whether you run a Shopify store, a WordPress blog, or a Webflow marketing site, this guide will walk you through the exact steps to create and deploy an llms.txt file-plus the strategic decisions that separate a useful file from a wasted one.

What llms.txt Actually Is (and What It Isn't)

Before you create anything, you need a clear mental model of what you're building. llms.txt is more like a curated sitemap.xml that includes only the very best content designed specifically for AI comprehension and citation. It is not a robots.txt replacement. It doesn't block crawlers, dictate indexing behavior, or restrict access to content.

The distinction matters because many articles conflate these files. Robots.txt is generally used to let automated tools know what access to a site is considered acceptable, such as for search indexing bots.

llms.txt information, on the other hand, will often be used on demand when a user explicitly requests information about a topic. Think of robots.txt as the bouncer at the door. Think of llms.txt as the concierge handing AI visitors a personalized guide to the building.

The expectation is that llms.txt will mainly be useful for inference-at the time a user is seeking assistance-as opposed to for training. When someone asks ChatGPT "what does Company X do?" or "which product should I choose?", the AI reaches for content it can parse quickly and trust. Your llms.txt tells it where to start. One honest caveat: llms.txt is currently just a proposed standard rather than something officially adopted. None of the LLM companies like OpenAI, Google, or Anthropic have officially said they're following these files when they crawl websites. Server log data from Semrush's test on Search Engine Land showed zero visits from Google-Extended bot, GPTbot, PerplexityBot, or ClaudeBot from mid-August to late October 2025.

So why bother? Two reasons. First, Anthropic, creator of Claude, specifically asked Mintlify to implement llms.txt and llms-full.txt for their documentation -a signal that at least one major AI company sees value in the format. Second, Vercel reports that 10% of their signups now come from ChatGPT, and Google included a llms.txt file in their new A2A protocol. The implementation cost is near zero. The potential upside is substantial.

The llms.txt Specification: Anatomy of the File

The llms.txt file is unusual in that it uses Markdown to structure the information rather than a classic structured format such as XML.

The reason is that many of these files will be read by language models and agents.

According to the official specification, a compliant file contains sections in a specific order:

• H1 heading with your project or site name.

This is the only required section.

• Blockquote with a short summary containing key information for understanding the rest of the file.
• Zero or more Markdown sections (paragraphs, lists) of any type except headings, with more detailed information.
• H2 sections organizing links to relevant resources, each with a URL and optional description.
• An "Optional" section specifically marking less critical resources.

Here's what a basic file looks like in practice:

# Your Company Name

> Brief, factual description of what your company does and who it serves.

Key context the AI needs to interpret your content correctly.

## Core Pages {#core-pages}

- [About Us](https://example.com/about): Company background, team, and mission
- [Products](https://example.com/products): Full product catalog with specifications
- [Pricing](https://example.com/pricing): Current plans and pricing structure

## Documentation {#documentation}

- [Getting Started](https://example.com/docs/start): Onboarding guide for new users
- [API Reference](https://example.com/docs/api): Complete API documentation

## Optional {#optional}

- [Blog](https://example.com/blog): Industry insights and company updates

Be specific. "Web hosting company" is too vague. "UK-based managed WordPress hosting provider with data centres in the UK, US and Asia" tells AI exactly what you offer and where.

llms.txt vs. llms-full.txt: Do You Need Both?

The specification defines two distinct files: /llms.txt, a streamlined view of your documentation navigation, and /llms-full.txt, a comprehensive file containing all your documentation in one place.

Where llms.txt points to sources, llms-full.txt contains the entire content of a website's documentation in a single Markdown file.

If llms.txt is the executive brief, llms-full.txt is the full-length book.

For most business websites, start with llms.txt. It covers the core use case: giving AI a curated map. Add llms-full.txt later if you have extensive documentation, an API reference, or a large product catalog where exact wording matters.

Step 1: Audit Your Content Before Writing a Single Line

Most guides skip this step. Don't. The quality of your llms.txt file depends entirely on which pages you include and exclude.

Start by selecting the pages you want AI systems to crawl first. Don't dump your whole sitemap into your llms.txt file.

AI systems have limited context windows. Every low-value page you include pushes something useful out.

Include these:

• Homepage and About page
• Core product or service pages
• Pricing (if public)
• Key documentation or help articles
• Your 3–5 best-performing blog posts
• Contact and location information

Exclude these: - Legal pages like privacy policy and terms of service-AI won't cite these, and they dilute your file.

• Outdated blog posts
• Duplicate pages covering the same topic
• Login pages, admin areas, checkout flows
• Aggressive sales copy

If your pricing is public, include it. AI tools frequently get asked "how much does X cost?" and they'll pull from your llms.txt if it's there. Without it, they guess, and guesses are often wrong.

Write descriptions in plain, factual language. Your llms.txt should feel more like a well-written instruction to a helpful assistant than a dense marketing deck. Avoid abstract descriptions, jargon, or buzzwords.

Step 2: Create the File Using the Right Tools

You can write your llms.txt manually in any text editor-VS Code, Sublime Text, or even Notepad. You need to format the file using Markdown. Save it as llms.txt (lowercase, no variation). If you'd rather not start from scratch, several free generators can create a starting template:

• Firecrawl's generator at llmstxt.firecrawl.dev:

automates the entire process, turning your website into structured files using gpt-4o-mini.

• Mintlify's generator:

paste in your docs site URL, and it will generate a starter file based on your structure.

• WordLift's generator and SiteSpeakAI's generator both offer free options.

These generators produce serviceable starting points. But treat the output as a draft, not a finished product. You can customize section names, add human-readable descriptions, and reduce URLs to just the most useful and representative ones.

Make sure the file is named exactly llms.txt. Variants like llm.txt or LLMS.txt will not work. Filenames are case-sensitive.

WordPress: Three Paths from Simple to Advanced

WordPress gives you the most options for implementation. Pick the path that matches your comfort level.

Option A: Yoast SEO (Fastest)

Generating your llms.txt file is included for free. No setup is required; just enable it in your site feature settings.

Here's the exact workflow: In your WordPress Dashboard, go to Yoast SEO → Settings → Site Features. Under AI tools, go to LLMS.txt and toggle the switch to on.

Click Save changes.

For more control, go to Customize llms.txt file, select Manual page selection, choose and add the pages you want, and click Save changes.

One limitation to know: Yoast SEO includes the 5 latest updated posts/pages/custom post types in the llms.txt file, giving priority to cornerstone content first. A user review from the WordPress plugin directory noted that Yoast's LLMs.txt was extremely simplified and lacked depth, listing only ~5 posts and ~5 pages. If your site has extensive content, you may want a more comprehensive option.

Option B: Dedicated Plugins (More Control)

The Website LLMs.txt plugin automatically generates and manages LLMS.txt files with full Yoast SEO, Rank Math, SEOPress, and AIOSEO integration. It offers features Yoast's built-in option lacks: choose post types, customize file titles or descriptions, attach optional Markdown files, and trigger manual regeneration.

AIOSEO also has built-in support. Go to AIOSEO → General Settings, click on the Advanced tab, and you'll find the "Generate an LLMs.txt file" option-usually enabled by default.

Option C: Manual Upload (Full Control)

For total control over every line, create the file yourself and upload it via FTP or your hosting file manager. Install a file manager plugin or access your site via FTP/SFTP. Navigate to the root directory, create a new file named llms.txt, paste the content, and verify access at yoursite.com/llms.txt.

This approach works best for sites where you want to handcraft every description and link-particularly if you have a content strategy that doesn't map neatly to WordPress post types.

Shopify: Working Around Platform Constraints

Unlike WordPress, Shopify doesn't give you direct access to the "root" directory of your store. You can't just upload a file there. This creates a genuine implementation challenge that requires a workaround.

The CDN + Redirect Method

Instead, upload the file to Shopify's content delivery network and then tell the world where to find it using a redirect. Here's the process: 1. Create your llms.txt file using a generator or text editor. 2. Upload your llms.txt file to Shopify's file manager (Settings > Files), copy the assigned Shopify CDN URL.

3. Create a URL redirect in Online Store > Navigation that redirects from /llms.txt to the CDN URL.

Be aware of the limitation reviewers have flagged: the llms.txt standard specifies that the file should be at the root of your domain. Since this is not possible directly, you need to create a redirect, which is not ideal but is the only option.

Shopify Apps

Several apps now automate this process. The LLMs.txt Generator app helps merchants easily create an llms.txt file that defines how AI search engines interact with their Shopify store. Other options include Arc's llms.txt generator and Revhope's AI Search Optimizer.

While the CDN redirect works, using a dedicated app is generally more reliable and easier to maintain. Apps also handle automatic updates when your product catalog changes-a significant advantage for stores with dynamic inventory.

The Liquid Template Approach

For more technical Shopify users, creating a page with the handle "llms" and assigning a custom Liquid template provides immediate llms.txt functionality, dynamically pulling current product information, collections, and store policies. This method keeps your file in sync with your catalog without manual updates.

Webflow: Native Support That's Refreshingly Simple

Webflow has made llms.txt implementation straightforward. With Webflow's latest update, you no longer need complex workarounds. You can now upload your LLMs.txt file directly through the project settings.

The process: 1. Create your llms.txt file in a text editor. 2. Upload it within your site's SEO settings.

3. Webflow will make the llms.txt file available in your domain's root, while also ensuring it does not get indexed by search engines.

4. Publish your site- Webflow requires you to re-publish after uploading. If you forget this step, the file won't be served live.

One important restriction: llms.txt upload is only available for CMS and Business plans. Starter or Basic plans will not expose the file even if uploaded.

After publishing, verify your file is accessible by visiting yourdomain.com/llms.txt in a browser. The Markdown content should display as plain text.

Content Strategy: What Separates a Good File from a Great One

A technically correct llms.txt file that points to the wrong content is worse than no file at all. Here's how to think strategically about what goes inside.

Write for Inference, Not Indexing

llms.txt doesn't contribute content to the model's memory; it simply tells the model where to look while it's actively generating a response. This means your file needs to surface content that answers the questions people actually ask AI about your brand, products, or industry. Ask yourself: If someone asked ChatGPT "What does [your company] do?" or "Is [your product] worth buying?", which pages would produce the best answer? Those pages belong in your llms.txt.

Prioritize Factual Density Over Marketing Language

AI models can detect-and tend to ignore-pure promotional copy. AI systems can spot aggressive marketing language and tend to ignore it. Your product page that lists specifications, comparisons, and concrete use cases is far more valuable than a landing page filled with superlatives. Include pages with data: case studies with numbers, pricing tables, feature comparisons, technical specifications. These are the pages AI systems are most likely to cite accurately.

Update on a Schedule

Review and update your LLMs.txt file quarterly or whenever you make significant changes to your site structure, launch new services, or add important content. An outdated llms.txt is worse than a stale sitemap because AI models that encounter deprecated URLs may generate inaccurate responses about your business. If you're using a WordPress plugin or Shopify app that auto-generates the file, verify the output periodically. Automated tools can't make strategic decisions about which content best represents your brand.

The Honest Assessment: Expectations and Limitations

Let's be straightforward about where things stand. As of 2026, llms.txt has gained moderate attention in the web and AI communities. Adoption is not yet universal, and while several leading AI firms respect llms.txt, compliance across the industry remains inconsistent.

We can see that llms.txt adoption has continued to rise, but according to Search Engine Land, 8 out of 9 sites saw no measurable change in traffic after implementation. John Mueller's comparison to the keywords meta tag remains the strongest skeptical argument: Mueller raises the point that an LLMs.txt file is redundant because why use that markdown file if the original content and structured data have already been downloaded?

The counterargument from practitioners is practical: there might be no observed benefit, but there's also no risk. Implementation takes 30 minutes. You're not betting resources on an uncertain outcome-you're placing a low-cost option on a potential future standard. Beyond the file itself, your broader GEO strategy should include ensuring your robots.txt doesn't accidentally block AI crawlers, maintaining clean structured data markup, and creating content that answers questions factually rather than persuasively.

Verification and Maintenance

After deploying your llms.txt file, run through this checklist: 1. Accessibility check: Visit yourdomain.com/llms.txt in your browser. The content should render as plain text. 2. Markdown validation: Paste your file into a Markdown previewer to catch formatting errors. 3. Link audit: Click every URL in the file. Broken links are worse than missing pages. 4. Content review: Read each description as if you were an AI system with no prior knowledge of your brand. Does it make sense? 5. Monitoring: Some plugins offer AI crawler detection-opt in to track whether GPTBot, ClaudeBot, or PerplexityBot are actually reading your file.

For ongoing maintenance, tie your llms.txt updates to your content calendar. When you publish a major new page, launch a product, or retire old content, reflect that change in the file. --- The llms.txt standard sits at an interesting inflection point. It's not universally adopted. Major AI providers haven't formally committed to it. And the evidence for direct traffic impact is thin. But creation and adoption of llms.txt is one of the few signs of LLM-first decision-making that website owners can make today.

Mintlify originally developed llms-full.txt in a collaboration with Anthropic, who needed a cleaner way to feed their entire documentation into LLMs without parsing HTML. Google included the format in their A2A protocol. These aren't fringe signals. The implementation math is simple. Thirty minutes of work. Zero risk. A structured signal to AI systems that your content is maintained, curated, and ready for machine consumption. Whether that signal pays dividends in six months or two years, the cost of placing the bet is as close to zero as you'll find in any optimization practice.