There are now over 15 Shopify apps that generate llms.txt files. Some are free, some charge $15/month, and one charges $499/month for “enterprise llms.txt optimization.” The marketing promises are bold: “Get found by ChatGPT,” “AI-ready product feed,” “Optimize for LLMs.”

We built llms.txt generation into our own app. And we think you deserve an honest look at what this file actually does today — and what it doesn’t.

What Is llms.txt?

The llms.txt spec proposes a simple idea: a Markdown file at /llms.txt on your website that serves as a machine-readable summary of your content. Think of it as robots.txt for AI — instead of telling crawlers what NOT to index, it tells them what IS worth reading.

For e-commerce, an llms.txt file typically lists your products grouped by category, with links, metadata (color, material, attributes), and descriptions. The format uses standard Markdown: H1 for the site name, H2 for sections, and bullet points with links.

The concept is clean. A structured, lightweight file that any AI system could parse to understand your product catalog without crawling every page.

The Promise vs. The Reality

Here’s where it gets uncomfortable.

No major AI crawler currently requests or reads llms.txt. This isn’t speculation — it’s observable data:

• Google’s John Mueller has publicly stated that Google does not use llms.txt.
• CDN log audits across multiple high-traffic e-commerce sites show zero requests for /llms.txt from GPTBot (ChatGPT), ClaudeBot (Anthropic), PerplexityBot, or Google-Extended.
• The llmstxt.org spec itself doesn’t claim adoption by major AI providers. It’s a proposal, not a standard that has been implemented by the platforms merchants care about.
• OpenAI’s official documentation for web search and shopping does not mention llms.txt as an input source. ChatGPT Search uses Bing’s index. ChatGPT Shopping uses merchant feeds via partnerships (like Shopify’s native integration).

To be clear: we’re not saying the concept is bad. We’re saying the file isn’t being read by the AI engines that matter for e-commerce today.

Why 15+ Shopify Apps Still Sell It

If llms.txt doesn’t work yet, why is there an entire app category built around it?

Three reasons:

1. It’s easy to build. Generating a text file from product data takes a few hours of development. Compare that to building JSON-LD injection, IndexNow integration, or AI image analysis — those are weeks of work.
2. The narrative is compelling. “A file that tells ChatGPT about your products” is an easy story to sell. Merchants don’t check CDN logs.
3. Future-proofing is a valid argument. The spec might get adopted. Having the file ready costs nothing. It’s a defensible “why not?” feature.

The problem isn’t that these apps exist. It’s that some position llms.txt as their primary (or only) AEO feature, which gives merchants a false sense of security.

Should You Still Have One?

Yes — but with the right expectations.

Having a well-structured llms.txt file is a low-cost insurance policy. If the spec gains adoption in the future, you’re already set. Some smaller AI tools and research crawlers do read it. And it’s a nice, human-readable summary of your catalog that could have secondary uses (developer documentation, partner integrations).

What you should NOT do:

• Pay $15+/month for an app whose only feature is llms.txt generation
• Consider your AEO strategy “done” after adding llms.txt
• Use llms.txt adoption as your primary metric for AI visibility

What Actually Works Instead

If llms.txt isn’t moving the needle today, what is? Based on documented mechanisms and observable data:

1. Schema.org JSON-LD (High Impact)

This is the structured data format that Bing, Google, and every major search engine actually reads. A rich Product JSON-LD block on your product pages — with category, color, material, GTIN, reviews, price, availability — is the single most impactful thing you can do for AI visibility.

When ChatGPT Search pulls product information, it comes from Bing’s index. Bing’s index is built from crawling your pages and reading your structured data. This chain is proven and documented.

2. IndexNow (High Impact)

The IndexNow protocol notifies Bing and Yandex that a URL has been updated, triggering a re-crawl within minutes to hours. The chain: IndexNow notification to Bing, Bing re-crawls your page, Bing updates its index, ChatGPT Search uses the updated index.

This is the fastest way to get product changes reflected in AI search results.

3. GTIN/EAN/UPC in Structured Data (High Impact for Shopping)

ChatGPT Shopping and Perplexity Shopping use GTINs to identify and match products. Without a GTIN in your structured data, AI shopping platforms rely on fuzzy name matching — which is unreliable.

4. Product Data Quality (Medium Impact, Compounds Over Time)

AI engines prefer products with specific, factual attributes over generic marketing copy. “Navy blue merino wool crew-neck sweater, 250gsm” gives an AI engine far more to work with than “Our bestselling must-have sweater!“

5. Shopify Agentic Storefronts (Emerging)

Shopify has launched native catalog syndication to ChatGPT, Perplexity, and Microsoft Copilot. This is happening at the platform level — your product data in Shopify Admin is being made available to AI engines directly. This will likely become the dominant channel for Shopify merchants.

The Honest Recommendation

Build the full pipeline. Don’t rely on a single file.

Your AEO strategy should look like this, in priority order:

1. Rich JSON-LD on every product page (with GTIN)
2. IndexNow notifications after product updates
3. High-quality, structured product attributes
4. llms.txt as a spec-compliant bonus

This is exactly why we built Agentic Flow as a complete pipeline rather than a single-feature tool. It does generate llms.txt — but it also scans product images with Google Gemini to create structured metadata, injects Schema.org JSON-LD automatically, and sends IndexNow notifications after every scan. Because a text file alone isn’t a strategy.

The merchants who will succeed in AI search are the ones who focus on the mechanisms that are proven today, while keeping a door open for the standards that might matter tomorrow.

---

Want the full picture? Read our guide to getting your products into ChatGPT or our complete AEO guide for Shopify merchants.