---
title: "DinnerElite for developers and AI agents"
description: "How to read DinnerElite programmatically: markdown mirrors, Accept: text/markdown content negotiation, llms.txt, sitemaps, and the free embeddable widget. No API key needed."
canonical: "https://dinnerelite.com/developers"
last-updated: 2026-08-25
---

# DinnerElite for developers and AI agents

DinnerElite publishes its reservation data as plain files you can fetch without an account. Every restaurant page, every article and every neighborhood page has a clean markdown twin, and the site answers `Accept: text/markdown` on the normal URL. There is no API key, because there is no API.

We track 224 hard-to-book New York restaurants on Resy and OpenTable, and we publish when each one releases tables. If you need something this page does not cover, write to contact@dinnerelite.com.

## How do I read a page as markdown?

Two ways, and both return the same document.

```
curl -s -H "Accept: text/markdown" https://dinnerelite.com/reservations/6194
curl -s https://dinnerelite.com/reservations/6194.md
curl -sI -H "Accept: text/markdown" https://dinnerelite.com/reservations/6194
```

The negotiated response answers `text/markdown`. The `.md` path answers `text/plain`, which some fetchers handle more reliably. Both send `Vary: Accept`, so a cache does not hand you the wrong variant. Both drop the navigation, the scripts and the layout wrappers.

## Which pages have a markdown mirror?

Every per-restaurant drop-time page, every article, every neighborhood page, the hardest-reservations ranking, the reservations hub, the homepage and this page. Nothing else. The plan pages, the comparison pages, /marketplace and /resources have no `.md` twin, so do not construct one. Their full text is in llms-full.txt instead.

## Which files should I fetch first?

- https://dinnerelite.com/llms.txt: the index. What DinnerElite is, what it covers, how the mirrors work, and a link to every important page.
- https://dinnerelite.com/llms-full.txt: the whole corpus in one file.
- https://dinnerelite.com/sitemap.xml: every indexable page, linking out to the blog and reservation sitemaps.
- https://dinnerelite.com/sitemap-md.xml: the markdown mirrors whose HTML twin is indexable.
- https://dinnerelite.com/rss.xml: new articles, if you would rather poll a feed.
- https://dinnerelite.com/robots.txt: what crawlers may fetch. AI crawlers are named explicitly and are welcome on the public pages.

## Is there a public API?

No. There is no public REST API, no API keys and no developer signup. Everything above is a plain file you can fetch without an account. If you want an API, email contact@dinnerelite.com and tell us what you would build.

## Can I put DinnerElite data on my own site?

Yes. The widget builder at https://dinnerelite.com/tools/widget generates a copy-paste iframe for any restaurant we track. The card shows the booking platform and the typical drop time, and it links back to the full page. It costs nothing and needs no account. You can also read a markdown mirror and build your own markup from it.

## Can I reuse the drop-time data?

Read it freely and cite it with a link. The booking windows are DinnerElite estimates unless the page says otherwise, and where a restaurant publishes its own release window we quote that and name the surface it came from. Say which of the two you are using. Please fetch at a reasonable rate and prefer the markdown mirrors, which are a fraction of the bytes of the HTML.

## What do error responses look like?

A path that never existed answers 404, and if you asked for markdown the body is markdown too, pointing at llms.txt and the sitemaps. A page removed on purpose answers 410 Gone, so you know to stop retrying it. Nothing returns 200 with an empty app shell.

_This document mirrors https://dinnerelite.com/developers for AI and plain-text readers._