PrestaShop and AI Search Engines: llms.txt and Lightweight Images

  • Home
  • PrestaShop and AI Search Engines: llms.txt and Lightweight Images
PrestaShop and AI Search Engines: llms.txt and Lightweight Images

Most of what gets a PrestaShop store ready for AI search engines is the same work that helps regular SEO: fast product pages and a clear description of what you sell. An llms.txt file and lightweight images are two cheap steps you can cover with ready-made modules in a single afternoon. Below we explain what each one actually does for PrestaShop and AI search engines, how to set them up step by step, and at what point a store owner needs a developer rather than another module.

What do AI search engines take from an online store?

AI assistants and AI answers in search engines use whatever they can fetch and understand: product pages, categories, and shipping and returns information. If Google does not index a page or the page takes too long to load, no extra trick will make up for it.

This is worth saying plainly before anyone promises a magic file. In its guide to optimizing for generative search, Google says you do not need llms.txt or any other special files to appear in Google, including its AI features, because Google Search does not use them. The llms.txt file is a proposed standard for other tools built on language models, described at llmstxt.org. So treat it as a cheap add-on, not as the foundation of your visibility.

We cover the format itself, its limits and the skepticism around it in more detail in our article on what the llms.txt file is. Here we focus on how it works in practice in PrestaShop.

llms.txt in PrestaShop: a ready-made module or manual work?

For most stores a ready-made module is enough, because llms.txt is a simple text file in the root directory that needs to stay up to date. Writing it by hand and uploading it over FTP works once, and then the file goes stale as your categories change.

This is where the WS LLMS module for PrestaShop fits in. According to the module description and documentation, it works like this:

  • in the configuration you write the main content in a Markdown editor, meaning a description of the store, its offer and its policies,
  • you pick categories from the native PrestaShop category tree, and if you select none, the module takes all of them,
  • you see a live preview of the finished file before you save it,
  • the generate button saves the llms.txt file directly in the store’s root directory,
  • the file can be refreshed automatically once a day with a script run from cron.

The documentation lists PrestaShop 1.7 or newer and write access to the store’s root directory as requirements. The latter is sometimes blocked on certain hosting plans, so check it before installing.

Lightweight product images: why do they matter more than llms.txt?

Images are usually the heaviest part of a product page, and a slow page loses both customers and search visibility. Speed is worth working on regardless of AI, because it is a basic part of store quality that customers feel on every visit.

PrestaShop creates several size variants for every uploaded image, so an uncompressed original multiplies on disk and in transfer. Editing each file by hand before upload only works until the first employee forgets to do it.

The Compressify module for PrestaShop handles this through the Compressify.eu service API. From the module description and documentation:

  • it compresses product images automatically after upload, once the size variants are generated,
  • it lets you bulk compress images added earlier and retry files that failed,
  • it supports JPG, PNG, GIF and WEBP,
  • it can keep a copy of the original, show a before and after comparison and restore the original,
  • it has a compression quality setting and a retention period for backups.

You need a Compressify.eu account and an API key, and the module panel shows your plan status and the number of optimizations left.

Example: a ceramics store prepared in one afternoon

Take a PrestaShop store selling everyday ceramics, with a dozen or so categories and several hundred images uploaded over the years without any editing. The plan for getting PrestaShop ready for AI search engines in a store like this looks as follows:

  1. Check in Search Console whether your product pages are indexed. If they are not, start there, not with files for AI.
  2. Install the compression module, enter the API key and turn on keeping originals.
  3. Run bulk compression on old images and look at the before and after comparison on a few products with fine textures, because that is where quality loss shows first.
  4. Install the llms.txt module and write a short text: what the store does, who it is for, and how shipping and returns work.
  5. Choose the categories that should go into the file, check the preview and generate the file.
  6. Add a cron job so the file refreshes itself after category changes.

Finally, measure the speed of a few product pages before and after the change. If images were the main bottleneck, you will see the difference right away. If you see no difference at all, the problem lies elsewhere.

PrestaShop and AI search engines: when is a ready-made module not enough?

Ready-made modules solve the typical problem of a typical store. Custom work is needed when the cause lies in the theme, the data or the integrations, not in a missing file or missing compression.

  • The theme loads originals instead of thumbnails. Compression will make the files smaller, but a theme that puts full-size images on the product list will still be slow. That is a fix in the theme code.
  • The images do not live in PrestaShop. When images come from a PIM system, a wholesaler or an external server, you need a process on the integration side.
  • Information about your offer is created outside the store. If product range descriptions, availability or B2B terms live in an ERP, the AI file and product pages have to be fed from that source.
  • The server is slow, not the images. Long server response times, heavy modules and no cache call for an audit, which we do as part of our work on online stores built on PrestaShop.

A simple rule from our software house practice: cheap, ready-made steps and measurement first, then a decision about code. Writing your own llms.txt generator or your own compression for a store with a standard catalog means spending budget on a problem that is already solved.

What else should you check on the product page?

Language models and people need the same thing: specifics in the page content. The product name, dimensions, material, shipping time and return policy should be text on the page, not just words on an image. You will find the full list of elements in our article on what a product page needs, and the post on preparing your site for AI answers gives the wider context.

Frequently asked questions

Will llms.txt improve my store’s ranking in Google?

You should not expect it to. Google says plainly that its search engine does not use llms.txt files. The file may help other AI tools understand your offer faster, and with a ready-made module it costs very little.

Will image compression reduce image quality?

With settings that are too aggressive, yes, which is why it is worth keeping the originals and comparing the before and after versions on a few products. You set the quality in the module configuration, and the original can be restored from the backup.

Will the modules work on an older version of PrestaShop?

According to their documentation, both modules require PrestaShop 1.7 or newer. On older versions it is better to plan a store upgrade first, because that is usually more important work than the file or compression itself.

If you want to get PrestaShop ready for AI search engines without overpaying, start with your images and an up-to-date description of your offer. You can handle both steps with ready-made modules, and you will find details on compression, backups and API configuration on the page of the Compressify module for PrestaShop.

Book a free consultation

Provide your phone number or schedule a meeting