llms.txt for a restaurant website: what it is and what to put in it
When your hours sit on Instagram, your menu is a photograph, and your ordering link is buried in the footer, an assistant has to assemble the same facts as a hungry customer.
Nexara, a restaurant order and operations platform built in Amman, Jordan, gives every storefront an llms.txt file: a short, readable guide to the restaurant and its key pages. It explains what llms.txt is, what to put in it, and how to connect accurate restaurant information to an ordering path.
Figures in this post are illustrative examples, not a client's actual results.
What is llms.txt, and should your restaurant have one?
llms.txt is a proposed convention for publishing a short Markdown guide to a website. Markdown is plain text with simple headings and links. The file usually sits at the root of the website, at /llms.txt, where a person or a compatible AI tool can open it directly.
For a restaurant, its useful job is to identify the business and point readers towards authoritative information. That means the branch location, current menu and ordering page, with enough context to distinguish your restaurant from another business with a similar name.
Yes, it is worth adding when your website basics are in order and maintaining the file is a small task. If the menu still lists discontinued dishes or the phone number goes unanswered, fix those first. A concise guide inherits the accuracy of the pages it points to.
An AI assistant recommending lunch needs to establish what you serve and whether you are open. An assistant ordering on someone's behalf also needs a supported way to select items and complete checkout. Readable restaurant information and a working ordering path serve different parts of the customer's request.
What should go in a restaurant llms.txt file?
Start with the restaurant name as the main heading, followed by one factual sentence describing the business. Use the same spelling as the website and Google Business Profile. Include the Arabic name where customers use it, so both names clearly identify the same restaurant.
- Identity and cuisine: the public trading name, city, neighbourhood and the food you actually serve.
- Location and hours: the branch address, regular opening times and where to check holiday changes.
- Menu: a link to readable dishes, options, current prices and availability.
- Ordering: the direct ordering link, pickup or delivery choices, and where delivery coverage is checked.
- Contact: the public restaurant phone number, including the country code.
- Agent access: links supplied by your ordering provider for supported agent discovery and ordering.
Keep changing details close to their source. Linking to the current menu is easier to maintain than copying every dish and price into this file. For several branches, identify each location separately and link to its hours and ordering details. One city-wide address cannot tell a customer which branch will prepare dinner.
What does an llms.txt example look like?
This compact file is a filled example for Bayt Al-Summaq, including its public contact details and verified restaurant pages.
# Bayt Al-Summaq > Bayt Al-Summaq is a Jordanian restaurant in Jabal Al-Weibdeh, Amman, serving home-style lunches and grilled dishes for pickup and delivery. Location: 12 Al-Rainbow Street, Jabal Al-Weibdeh, Amman, Jordan. Hours: Daily, 12:00 to 23:00, Amman local time. Check the contact page for holiday hours. Phone: +962 6 555 0184. Ordering: Choose pickup or enter your delivery address on our ordering website to check coverage and charges. ## Menu and ordering - Current menu: Dishes, options, prices in JOD and availability. - Order directly: Choose items and complete the restaurant's checkout. ## Branch information - Location, hours and contact: Branch address, public phone number and opening-hour updates.
The entity sentence names the restaurant, places it in an Amman neighbourhood and identifies its cuisine. The links lead to facts that change during normal service, including availability and delivery charges. For agent access, ask your ordering provider for the exact discovery links supported by your storefront.
How does it sit beside robots.txt and structured data?
These files and page annotations have different jobs. A restaurant website can use all of them. When discussing AI crawlers with whoever maintains your site, ask them to check access to the actual menu and branch pages as well as the llms.txt file.
| Website element | What it tells a reader or tool | Restaurant use |
|---|---|---|
| llms.txt | A short description and selected links, following a proposed convention. | Identify the restaurant and point to its menu and ordering path. |
| robots.txt | Crawling instructions for automated visitors that follow those rules. | Check whether relevant crawlers may fetch public restaurant pages. |
| Structured data | Labelled facts attached to a page, such as Restaurant, Menu and MenuItem information. | Make identity, menu details and availability readable in a consistent format. |
| Visible website pages | The information customers can read and act on. | Show accurate hours and prices, and provide a usable checkout. |
robots.txt controls crawling instructions; it provides no password protection for private information. Keep customer records and private contact details out of public files. For the menu and business labels, see structured data for a restaurant website.
Consistency is the practical test. If the contact page says the branch closes at 23:00 while its Google profile says midnight, llms.txt introduces another place that needs correction. Match the facts across listings, then maintain them when hours or locations change.
How does Nexara connect the file to restaurant ordering?
Nexara's main website and every Nexara restaurant storefront publish llms.txt. Storefront files are bilingual and link to agent endpoints. Storefronts also publish Restaurant, Menu and MenuItem structured data with availability, and permit AI crawlers to access public content.
Agent ordering is built into every Nexara restaurant storefront for agents that speak ACP (2026-04-17), UCP or MCP. These formats let an assistant read the live menu and carry out an order for a customer. The module is not yet deployed.
The built-in flow gives the agent a server-priced delivery quote based on the restaurant's menu and delivery coverage. An anonymous agent verifies the customer's Jordanian phone through an SMS code; trusted platforms complete with a key. Payment is cash on delivery or pay at the counter for pickup.
The completed order enters the kitchen queue, prints in the kitchen and appears on the order board with an AI agent badge. The restaurant sets the menu prices; the assistant uses the calculated total. Read the Nexara agent ordering guide and Nexara features for the supported storefront capabilities.
What should you check before publishing?
Give the file a named owner, whether that is you or the person who maintains the website. An opening-hours change should trigger a check of the website, public listings and llms.txt together.
- Confirm the entity sentence, branch address and cuisine against the visible website.
- Open every menu and ordering link on a phone without signing into an owner account.
- Check that dishes and prices are readable as text, including paid options.
- Match regular and holiday hours with your Google Business Profile, and review customer feedback for repeated factual confusion.
- Open /llms.txt on the restaurant's domain and verify that the published text matches the approved version.
For a wider review of how assistants identify a restaurant, use the restaurant AI discovery checklist. Your public identity, reviews and working ordering link all deserve attention alongside the file.
What to remember
- Use llms.txt to identify the restaurant and point to authoritative pages.
- Keep menu prices, hours and branch details consistent across public sources.
- Use robots.txt and structured data for their separate purposes.
- Check every public ordering and contact path when restaurant details change.
Questions people ask
Is llms.txt required for a restaurant website?
It is optional. Add it as a concise guide once your public menu, hours and ordering page are accurate.
Will adding llms.txt make AI assistants recommend my restaurant?
Each assistant decides which sources to retrieve and which businesses to recommend. Accurate public facts, useful menu information and consistent listings give it clearer material to assess.
Where do I put the llms.txt file?
Publish it at /llms.txt on your restaurant's website domain. Ask your website provider to make it readable without a login.
Should I copy my whole menu into llms.txt?
Link to the current readable menu and briefly describe what it contains. Copying every price creates another document to update whenever the menu changes.
What should I update when my restaurant menu changes?
Update the live menu first, then check the llms.txt links, prices, availability and ordering path. Keep the entity, phone number, hours and branch details aligned across public listings.
Does Google use llms.txt for its AI results?
llms.txt is a proposed convention. Each assistant or search service decides which sources it uses. Keep your Google Business Profile, visible website, structured data and ordering path accurate alongside the file.