What Is llms.txt and Should Your SaaS Company Have One?

If you’re running a SaaS company and wondering whether llms.txt matters, here’s the short answer: it’s a simple text file that tells AI models what your company does, what you sell, and where to find key information. Think of it as a structured cheat sheet for large language models. The concept is still young, and the evidence for its impact is mixed, but the cost of creating one is close to zero. For B2B SaaS companies that rely on being found during buyer research, ignoring how AI tools read your site feels like a risk you don’t need to take.

The question of whether your SaaS company should have an llms.txt file comes down to a practical trade-off: minimal effort versus potential upside in how AI assistants represent your product. We’ll break down what the file is, how to build one, what to put in it, and whether the data supports its value.

What is llms.txt?

The llms.txt file is a proposed standard, originally outlined by Jeremy Howard in late 2024, that gives AI models a structured, plain-text summary of a website’s content. It sits at the root of your domain (yourdomain.com/llms.txt) and provides information in a format that’s easy for language models to parse. Where robots.txt tells search engine crawlers what they can and can’t access, llms.txt tells AI systems what your site is actually about.

The file uses Markdown formatting and typically includes your company name, a brief description, links to key pages, and optional sections for documentation, pricing, and product details. It’s not a formal web standard governed by the W3C or IETF. It’s a community-driven proposal that’s gained traction because it solves a real problem: most websites are built for human browsers, not for AI systems trying to synthesise accurate answers about your product.

For SaaS companies, this matters because buyers increasingly use AI tools like ChatGPT, Perplexity, and Gemini to research software categories and compare vendors. If those tools can’t find clear, structured information about what you do, they’ll either skip you or get it wrong. The llms.txt file is one way to reduce that risk, sitting alongside your existing SEO and content strategy rather than replacing any of it.

What llms.txt does

The file serves a single purpose: it makes your website’s most important information easy for AI models to consume. Most SaaS websites are full of JavaScript-rendered pages, gated content, navigation menus, and dynamic elements that humans handle fine but AI crawlers struggle with. An llms.txt file strips all that away and presents the essentials in clean, readable text.

When an AI tool is asked “What does [your company] do?” or “Compare [your product] to [competitor],” it pulls from whatever sources it can find and trust. Those sources include your website, review sites, documentation, and third-party mentions. The llms.txt file gives the AI a direct, authoritative summary from you, the source. It doesn’t guarantee the AI will use it, but it increases the chance your own description shapes the answer.

There’s an important distinction here. The file doesn’t control what AI models say about you. It’s not a directive like robots.txt. It’s more like a press release for machines: you’re putting your best, most accurate information where AI tools can easily find it. Some early analysis suggests the file is less impactful than proponents hoped, but the logic behind making your content AI-readable is sound regardless of whether this specific file format becomes the long-term standard.

The file also complements your broader GEO (Generative Engine Optimisation) efforts. Teams that build consistent brand descriptions across the web, earn third-party citations, and maintain structured content give AI tools more to work with. The llms.txt file is one piece of that puzzle.

How to create an llms.txt file

Creating the file takes less than an hour. You don’t need a developer, though having one review the final placement is sensible.

1. Open a plain text editor. Don’t use Word or Google Docs, as they add hidden formatting. Use something like VS Code, Sublime Text, or even Notepad.
2. Start with a Markdown heading using your company name, followed by a one-line description of what your product does.
3. Add a “Links” section listing URLs to your most important pages: homepage, product pages, pricing, documentation, API reference, and key landing pages.
4. Include an “Optional” section for secondary resources like blog posts, case studies, changelog, or integration guides.
5. Save the file as llms.txt (not llms.md, not llms.html).
6. Upload it to the root directory of your domain so it’s accessible at yourdomain.com/llms.txt.
7. Test the URL in a browser to confirm it loads as plain text.

Some companies also create an llms-full.txt file that includes more detailed content, essentially a longer version with expanded descriptions and additional context. This is useful if your product is complex or you operate in a category where AI tools frequently give incomplete or inaccurate answers. The specification suggests keeping the core file concise while using the extended version for depth.

One common mistake is treating the file like a keyword-stuffed SEO page. Don’t do that. Write it like you’re explaining your product to a smart colleague who’s never seen it before. Clear, factual, and structured.

What to include in it

The content of your llms.txt file should reflect what you’d want an AI tool to say about your company if a potential buyer asked. For a B2B SaaS company, that typically means covering these areas:

• Company name and a one-sentence positioning statement
• Product description covering what it does, who it’s for, and what problem it solves
• Links to your homepage, product pages, and pricing page
• Links to API documentation or developer resources if you have a PLG motion
• Links to key comparison or “vs” pages if you’ve built them
• Links to your most important integration pages
• A brief note on your target market or ideal customer profile
• Links to trust signals like security certifications, compliance pages, or review profiles

Avoid including blog posts that aren’t directly relevant to understanding your product. The file should be a curated selection, not a sitemap. If you sell to multiple personas (say, developers and marketing leaders), you can organise links by audience to make the structure clearer.

For companies with a product-led motion, linking to your signup or free trial page makes sense. For sales-led companies, linking to your demo request or contact page is more appropriate. The file should reflect your actual buyer journey, not a generic template.

Keep the total file under 2,000 words for the core version. AI models have context windows, and a shorter, well-structured file is more likely to be fully processed than a sprawling one.

Does llms.txt actually affect AI visibility?

This is where honesty matters more than hype. The evidence is mixed, and anyone telling you llms.txt is a guaranteed path to AI visibility is overselling it.

One test of the file’s impact found no statistically significant improvement in how often a brand appeared in AI-generated answers after adding llms.txt. The AI tools tested didn’t consistently reference the file’s content, and some appeared to ignore it entirely. That’s a data point worth taking seriously.

On the other hand, the UK industry AI visibility scorecard shows significant variation in how well different sectors and companies appear in AI-generated results. Companies with strong, consistent content across multiple sources tend to perform better. The llms.txt file alone won’t fix poor visibility, but it can contribute to a broader programme of making your brand AI-readable.

The honest position is this: llms.txt is a low-cost signal, not a silver bullet. It works best as part of a wider GEO strategy that includes consistent entity descriptions, strong third-party citations, structured data on your website, and content that directly answers the questions buyers ask AI tools. At Gripped, we run GEO audits that assess how ChatGPT, Perplexity, Gemini, and Claude currently describe a company, what sources they pull from, and where competitors appear instead. The llms.txt file is one input to that process, not the whole strategy.

If you’re a SaaS company with a Lighthouse score above 90, solid structured data, and content built around buyer questions, adding llms.txt is a sensible next step. If your website fundamentals are weak, fix those first.

Should your SaaS company have one?

Yes, but with realistic expectations. The file takes minimal effort to create, costs nothing to maintain, and positions you for a future where AI-mediated discovery becomes a larger share of how buyers find and evaluate software.

For B2B SaaS companies between £2M and £20M ARR, the calculus is straightforward. Your buyers are already using AI tools to research vendors. A growing number of tech companies are adopting llms.txt as part of their digital presence, and while adoption doesn’t prove effectiveness, being absent while competitors are present creates an unnecessary gap.

The real question isn’t whether to create the file. It’s whether you treat it as a standalone tactic or integrate it into a proper GEO and content strategy. A well-maintained llms.txt file paired with strong SEO foundations, consistent brand messaging, and third-party authority signals gives AI tools the best chance of representing your product accurately.

If your marketing team is stretched thin, this is a one-time task that takes an hour. If you’re working with an agency, it should be part of their GEO or SEO programme. The practical guidance for implementing llms.txt is well-documented, and there’s no technical barrier to getting it done this week.

Don’t expect it to move your pipeline metrics on its own. Do expect it to be one more signal that helps AI tools understand what you sell, who you sell it to, and why it matters.

Common questions

Where does the llms.txt file go?

The file sits at the root of your domain, accessible at yourdomain.com/llms.txt. This mirrors the convention used by robots.txt and security.txt. If you use a CDN or have multiple subdomains, place it on your primary marketing domain. Some companies also place a version on their docs subdomain (docs.yourdomain.com/llms.txt) if their documentation is a major entry point for AI tools. The file must be publicly accessible with no authentication required. If it returns a 404 or requires a login, AI crawlers won’t find it.

Is llms.txt a ranking factor?

No. It’s not a Google ranking factor, and no major AI provider has confirmed that they prioritise content from llms.txt files in their responses. The file is a proposed community standard, not an official protocol recognised by OpenAI, Google, Anthropic, or any other AI company. That said, making your content easier for AI systems to parse is directionally useful. The file won’t hurt your SEO, it won’t conflict with your robots.txt, and it takes minutes to implement. Think of it as a hygiene factor rather than a competitive advantage: it won’t win the race, but not having one means you’re giving AI tools less to work with than competitors who do.

Getting started

The llms.txt file is a small, practical step in a much larger shift toward AI-mediated buyer research. For SaaS companies, the real work is building the kind of content, authority, and consistency that AI tools need to represent you accurately. The file itself is just the entry point.

If you’re a marketing leader at a B2B SaaS or tech company and you’re trying to figure out how AI visibility fits into your broader growth strategy, Gripped works exclusively with companies like yours. We focus on pipeline and revenue, not vanity metrics, and our GEO audits show exactly where AI tools are citing your competitors instead of you. Get your free growth audit to see where you stand.