Airbnb Scraper

Single Apify actor that combines three search modes for Airbnb listings:

• search-fast — fast scrape of listing cards from /api/v3/StaysSearch (price, rating, basic facts; no detail page hit).
• search-full — search-fast + full listing details merged in (StaysPdpSections per listing).
• urls — full details for an explicit list of https://www.airbnb.com/rooms/<id> URLs.

HTTP-only — no Playwright/Selenium/Puppeteer/nodriver in the production image. The actor talks to Airbnb's internal Niobe GraphQL persisted-query endpoint directly through curl_cffi(impersonate="chrome") so TLS/HTTP-2 fingerprints look like real Chrome.

Modes

Mode	Inputs	Output	Speed	Cost
search-fast	locationQueries[]	cards only	fastest	cheapest
search-full	locationQueries[]	cards + listing details	medium	per listing
urls	startUrls[]	listing details	medium	per listing

If mode is omitted, it's auto-detected: startUrls present → urls, else search-fast.

Quick start (Apify Console)

Default prefill runs out of the box and is sized to satisfy Apify's quality test:

{
  "mode": "search-fast",
  "locationQueries": ["Prague"],
  "maxItemsPerQuery": 5,
  "currency": "USD",
  "locale": "en"
}

Example: search-fast

{
  "mode": "search-fast",
  "locationQueries": ["Prague", "Berlin"],
  "checkIn": "2026-06-01",
  "checkOut": "2026-06-07",
  "adults": 2,
  "minPrice": 50,
  "maxPrice": 250,
  "roomType": "Entire home/apt",
  "maxItemsPerQuery": 100,
  "currency": "USD",
  "locale": "en"
}

Returns one item per listing with: id, url, name, coordinates, rating, bedrooms/beds/bathrooms, price (for the requested dates, including original price and discounts), images, badges.

Example: search-full

{
  "mode": "search-full",
  "locationQueries": ["Prague"],
  "checkIn": "2026-06-01",
  "checkOut": "2026-06-07",
  "adults": 2,
  "maxItemsPerQuery": 30,
  "includeReviews": true,
  "includeCalendar": true,
  "currency": "USD",
  "locale": "en"
}

Same as search-fast plus from the listing detail: propertyType, host (Superhost status, years hosting, rating count, profile photo, cohosts), amenities (categorised), highlights, description, sleepingArrangement, houseRules, safety, full image set, and optionally reviews and availabilityCalendar.

Example: urls

{
  "mode": "urls",
  "startUrls": [
    { "url": "https://www.airbnb.com/rooms/53997462" },
    { "url": "https://www.airbnb.com/rooms/20277576?adults=2&check_in=2026-06-01&check_out=2026-06-07" }
  ],
  "adults": 2,
  "checkIn": "2026-06-01",
  "checkOut": "2026-06-07",
  "includeCalendar": true,
  "currency": "USD",
  "locale": "en"
}

Output schema

Every item contains (null keys are dropped):

{
  "id": "53997462",
  "url": "https://www.airbnb.com/rooms/53997462",
  "name": "Cozy apartment at historical center",
  "propertyType": "Entire condo",                  // search-full / urls only
  "roomType": "Entire home/apt",                   // search-full / urls only
  "personCapacity": 2,                             // search-full / urls only
  "bedrooms": 1, "beds": 2, "bathrooms": 1,        // from search card
  "coordinates": { "lat": 40.6298, "lng": 22.9497 },
  "neighborhood": "Thessaloniki, Greece",          // search-full / urls only
  "rating": {
    "value": 4.95,
    "reviewsCount": 65,
    "categories": {                                // search-full / urls only
      "accuracy": 4.91, "checkin": 4.97, "cleanliness": 4.94,
      "communication": 5.0, "location": 4.98, "value": 4.95, "overall": 4.95
    }
  },
  "images": [{ "id": "...", "url": "https://a0.muscache.com/..." }],
  "imageCount": 46,                                // search-full / urls only
  "price": {
    "currency": "USD",
    "displayPrice": "$954",
    "originalPrice": "$1,400",
    "qualifier": "total",
    "breakdown": { /* full Airbnb price breakdown */ }
  },
  "badges": ["Superhost", "Guest favorite"],
  "amenities": [{ "category": "Bathroom", "items": [...] }],  // search-full / urls
  "highlights": [{ "type": "...", "headline": "...", "body": "..." }],
  "description": "...",                            // search-full / urls
  "descriptionLanguage": "en",
  "sleepingArrangement": [{ "title": "Bedroom", "subtitle": "1 queen bed" }],
  "houseRules": ["Check-in after 3:00 PM", "Checkout before 11:00 AM"],
  "safety": ["Carbon monoxide alarm", "Smoke alarm"],
  "host": {
    "name": "Alexia", "isSuperhost": true, "ratingCount": 65,
    "ratingAverage": 4.95, "yearsHosting": 4, "monthsHosting": 4,
    "profilePictureUrl": "...", "cohosts": [...]
  },
  "reviews": {                                     // only when includeReviews=true
    "total": 65,
    "fetched": 50,
    "reviews": [{ "id": "...", "rating": 5, "comments": "...", "createdAt": "...", "reviewer": {...} }],
    "tags": [...]
  },
  "availabilityCalendar": {                        // only when includeCalendar=true
    "constantMinNights": 2,
    "months": [{ "month": 6, "year": 2026, "days": [{ "date": "2026-06-01", "available": false, ... }] }]
  },
  "locale": "en",
  "currency": "USD",
  "sourceUrl": "https://www.airbnb.com/rooms/53997462"
}

Pass "rawOutput": true to embed the raw GraphQL responses under _raw.

Notes & limits

• Currency is restricted to USD and EUR. Anything else falls back to USD. Rubles aren't supported on purpose — Airbnb left the Russian market in 2022.
• Search ceiling is 270 listings per query (Airbnb returns at most 15 pages of 18). Use multiple locationQueries to scrape more.
• Empty input is handled gracefully: the actor exits with code 1 and a readable status message instead of a stack trace.
• Proxy: The default is { "useApifyProxy": true } (Apify picks an available group automatically). If your plan has DATACENTER access, set apifyProxyGroups: ["DATACENTER"] for the cheapest path -- it works fine for Airbnb because curl_cffi(impersonate="chrome") handles TLS fingerprinting. Otherwise use ["RESIDENTIAL"].

Throughput & maxConcurrency

maxConcurrency is a global cap on parallel HTTP requests to Airbnb -- every operation (search, listing details, calendar, reviews, BookIt) shares the same budget. Every listing in search-full fans out into 1-4 HTTP calls depending on which enrich flags are on, and the global semaphore makes sure they don't all burst at once.

Measured on a clean machine with no proxy, hitting the live Airbnb API, with search-full + includeReviews + includeCalendar (3-4 requests per listing) and 25 listings per query:

maxConcurrency	items/sec	notes
3	~1.0	safe even for fragile residential proxies
5	~1.5	good balance for typical residential plans
10	~2.6	comfortable for datacenter proxies
20	~3.7	aggressive; datacenter / no-proxy territory
30	~4.4	maxed out -- Airbnb's own latency dominates
50	~4.9	no extra gain beyond ~30

Picking the right value

Your proxy	Recommended maxConcurrency
No proxy (useApifyProxy: false)	20-30
Apify DATACENTER	10-20
Apify RESIDENTIAL	3-5

If the log starts spamming network/proxy error (ProxyError / Timeout) or transient status=503 -- halve maxConcurrency and rerun. The actor has three independent retry budgets (proxy_failures, captcha_failures, attempt), so transient errors don't lose listings; they just slow the run down.

Expected wall time for a full run

A full 270-listing search-full + reviews + calendar query is ~810-1080 HTTP calls. At measured rates:

• DATACENTER, maxConcurrency=15: ~2 min
• RESIDENTIAL, maxConcurrency=5: ~10-15 min
• No proxy, maxConcurrency=30: ~1 min (but the IP gets hot fast)

Tech

• apify >= 3.3, < 4 SDK, crawlee >= 1.5, < 2
• curl_cffi >= 0.7.0 — Chrome TLS/HTTP-2 fingerprinting
• pydantic >= 2.7 — input validation
• Default container: 512 MB RAM (HTTP-only — no browser needed)