OpenTable + Resy Scraper

A single Apify actor that scrapes restaurant data from both OpenTable and Resy. Switch between platforms with the platform input. Find restaurants by city or cuisine, look up specific venues by URL or slug, and (on Resy) check live reservation availability.

No login required for either platform.

Features

• Two platforms in one actor — platform input picks OpenTable or Resy.
• Five modes per platform: search, by-restaurant, by-cuisine, by-city, by-URL.
• Rich record fields — name, cuisine, price range, neighborhood, address, coordinates, rating, photos, contact info, and full-text reviews (OpenTable detail pages).
• Resy availability (optional) — append live reservation slots from Resy's /4/find API.
• Cuisine taxonomy — 50-entry curated dropdown maps to OpenTable + Resy search terms.
• Country filter — drop records outside a target ISO country.
• Price-range filter — $ through $$$$, mapped to each platform's price-band IDs.
• Auto-escalation — TLS-impersonated curl_cffi first; auto-switches to Apify Proxy (datacenter, then residential) on 403 / 429.

Sample input

{
  "platform": "opentable",
  "mode": "search",
  "city": "New York",
  "cuisine": "italian",
  "partySize": 2,
  "maxItems": 10
}

{
  "platform": "resy",
  "mode": "byRestaurant",
  "urls": ["https://resy.com/cities/ny/venues/carbone"],
  "includeAvailability": true,
  "day": "2026-07-15"
}

Modes

Mode	OpenTable	Resy
search	/s/?term=...&metroId=... — extracts multiSearch.restaurants[] from the SSR window-vars JSON.	POST /3/venuesearch/search with geo + query.
byRestaurant / byUrl	Detail page /{slug} — extracts restaurantProfile.restaurant + reviewsData.	GET /3/venue?url_slug=...&location=... — full venue detail.
byCuisine	Search with cuisine as the search term.	Search with cuisine as the query.
byCity	Search with city / metroId only.	Search by geo coordinates.

Output (per platform)

Both platforms expose a unified entityId (string) for cross-platform joins. The platform-specific restaurantId (OpenTable) and venueId (Resy) are still emitted alongside entityId for backwards compatibility.

OpenTable

{
  "platform": "opentable",
  "entityId": "96577",
  "restaurantId": "96577",
  "name": "Sistina",
  "url": "https://www.opentable.com/sistina",
  "primaryCuisine": "Italian",
  "priceBand": "$31 to $50",
  "priceBandId": 3,
  "priceLabel": "$$$",
  "currencySymbol": "$",
  "neighborhood": "Upper East Side",
  "address": {"line1": "24 E 81st Street", "city": "New York", "state": "NY", "postCode": "10028", "country": "United States"},
  "coordinates": {"latitude": 40.776, "longitude": -73.962},
  "totalReviews": 1894,
  "overallRating": 4.7,
  "isInstantBookable": true,
  "hasPrivateDining": true,
  "coverImageUrl": "https://resizer.otstatic.com/...",
  "recordType": "restaurant",
  "scrapedAt": "..."
}

In byRestaurant mode the record additionally includes recentReviews[], hoursOfOperation[], cuisines[], contact, executiveChef, dressCode, paymentOptions, parkingDetails, etc.

Resy

{
  "platform": "resy",
  "entityId": "6194",
  "venueId": 6194,
  "name": "Carbone",
  "urlSlug": "carbone",
  "url": "https://resy.com/cities/ny/venues/carbone",
  "cuisine": ["Italian"],
  "neighborhood": "Greenwich Village",
  "priceRangeId": 4,
  "priceLabel": "$$$$",
  "currencySymbol": "$",
  "coordinates": {"latitude": 40.728, "longitude": -74.000},
  "rating": {"average": 4.74, "count": 31498},
  "address": {"line1": "181 Thompson Street", "city": "New York", "state": "NY", "country": "United States", "countryIso": "US"},
  "isGlobalDiningAccess": true,
  "coverImageUrl": "https://image.resy.com/...",
  "recordType": "restaurant",
  "scrapedAt": "..."
}

With includeAvailability=true (byRestaurant mode), each Resy record gains:

"availability": {
  "day": "2026-07-15",
  "partySize": 2,
  "slotCount": 6,
  "slots": [{"start": "...", "type": "Dining Room", "id": "..."}]
}

High-churn output fields

Some fields evolve rapidly between runs and should be treated as snapshot-in-time:

• Resy availability.slots[] and availability.slotCount — Resy slots change every few minutes as bookings fill up; the same (venue, day, partySize) query can return different counts on consecutive runs. Compare across the timestamp captured in scrapedAt, not against historical exports.
• OpenTable recentReservationCount — daily booking counter; rolls over each midnight in the venue's local timezone.
• overallRating, totalReviews, Resy rating.average/rating.count — providers' aggregate review stats update as new reviews are posted; expect minor drift on each run.
• recentReviews[] (OpenTable) — most-recent review window changes daily; absent from a run does not mean the venue lost reviews.
• Photos and coverImageUrl — CDN URLs may rotate; the resource itself is usually stable but the URL string can change.

Stable fields safe for diff/dedup: restaurantId/venueId, entityId, name, url, address, coordinates, priceBand/priceRangeId, cuisines/primaryCuisine.

Limitations

• Resy reviews are not exposed in the public API — only aggregate rating + count. Use OpenTable for full review text.
• OpenTable detail URL has two forms: /{slug} (legacy) and /r/{slug} (newer). Only one form returns a populated SSR page per restaurant; the actor probes the user-supplied form first, then auto-falls back to the alternate.
• OpenTable search returns ~50 restaurants per page — beyond that, paginate by changing metroId or cuisine.
• Resy availability requires a day parameter — the actor defaults to 14 days from today; override via the day input. Dates farther than 365 days in the future are rejected.
• The Resy widget api_key may rotate; the actor harvests a fresh one from resy.com/modules/app.*.js at startup if the bundled key fails.

FAQ

Do I need login cookies? No. Both APIs are public, served from datacenter IPs.

Does it support international markets? Yes — OpenTable covers 25+ countries (US, CA, UK, France, Germany, Italy, Spain, Australia, Mexico, etc.). Resy covers US, UK, Mexico, Australia, France, and a handful of other markets. Use the country filter to scope output.

Can I scrape both platforms in one run? Not in a single call — run the actor twice with platform=opentable and platform=resy and merge the datasets downstream.

Why does my run keep getting 403? OpenTable is fronted by Akamai BMP. The actor auto-escalates to Apify Proxy (datacenter, then residential) on 403/429 if autoEscalateOnBlock=true.

Is this legal? This actor only fetches public, unauthenticated restaurant directory data (names, addresses, cuisines, public reviews) — the same data any visitor sees on the website. No reservations are placed and no user data is collected. You are responsible for complying with each platform's Terms of Service in your jurisdiction.