Craigslist Listings Scraper (Apify Actor)

Crawl Craigslist listings by city + category + optional free-text query. Returns one dataset record per result with title, price, location, post URL, ISO 8601 posted-at timestamp, first image URL, and (for the top results per page) a 500-char excerpt of the post body. No auth, no proxy, no captcha. Reads public static HTML + JSON-LD.

Source coverage

Source	Auth	Captcha	Cities	Categories	Use
https://{city}.craigslist.org/search/{category}	none	none	~700 city subdomains (newyork, losangeles, sfbay, chicago, boston, seattle, etc.)	sss (for sale), cta (cars+trucks), roo (rooms/shared), apa (apartments), etc.	browse + search results
https://{city}.craigslist.org/view/d/{slug}/{id}	none	none	same	per listing	post body + timestamp

All endpoints return HTML + JSON-LD, no proxy / captcha / JS rendering needed, no auth, no API key. Craigslist serves standard HTML to anything that looks like Chrome; the actor sets a Chrome User-Agent automatically.

What it does

• Crawl by city + category + optional searchQuery. Pagination is ?s=<offset> where offset = 120 * pageIndex.
• For each search-result page, the actor:
  1. Pulls <script id="ld_searchpage_results"> JSON (rich structured data: title, price, neighborhood, geo, image).
  2. Pulls all /view/d/.../... hrefs in result order from the DOM.
  3. Pairs them by index (itemListElement[i] ↔ hrefs[i]). Items without a matching href are skipped (CL occasionally trims; the actor never hallucinates URLs).
  4. Pushes one Dataset record per item.
  5. If fetchBodies=true, follows the first 5 post URLs per page to grab the post body + timestamp.
• Detail fetcher pulls #postingbody text (strips the "QR Code Link to This Post" boilerplate), trims to 500 chars, and pulls <time.datetime> as the ISO posted-at timestamp.
• Pairs JSON-LD with DOM hrefs by position (not by slug match) for resilience to markup changes — see the "Why JSON-LD + DOM-order pairing" section in src/main.js.

Input

Field	Type	Default	Description
city	textfield	"newyork"	Craigslist city subdomain, lowercase, no spaces — e.g. newyork, losangeles, sfbay, chicago, boston, seattle. See craigslist.org/about/sites for the full ~700-city list.
category	textfield	"sss"	3-letter category code. sss = all for sale (default), cta = cars+trucks, roo = rooms/shared, apa = apartments, sub = sublets. For sub-categories like cta?auto_make_model=..., pass the full query string — the actor URL-encodes it. Empty = browse all categories.
searchQuery	textfield	""	Free-text query appended to the search URL. Empty = browse the category.
maxPages	number (1-25)	3	Number of search-result pages to crawl. Each page has ~120 results.
fetchBodies	checkbox	true	If true, fetches the first 5 post bodies per page to capture the post body excerpt and posted-at timestamp. Adds 1 detail request per item (so 5 extra requests per page).

Output

One dataset record per listing. See .actor/dataset_schema.json for the full shape. Sample (first record from a city=sfbay, category=cta, searchQuery=bicycle, maxPages=2, fetchBodies=true smoke run, abridged):

{
  "city": "sfbay",
  "category": "cta",
  "searchQuery": "bicycle",
  "title": "Specialized road bike — size 56cm",
  "price": "$450",
  "priceValue": 450,
  "location": "San Francisco",
  "url": "https://sfbay.craigslist.org/pen/d/specialized-road-bike-size-56cm/1234567890.html",
  "postedAt": "2026-07-08T14:23:18-0700",
  "bodyExcerpt": "Selling my Specialized Allez road bike, size 56cm. Lightly used, new tires in May. Includes pedals and a bike computer. Cash only, no trades. Pickup in the Mission.",
  "image": "https://images.craigslist.org/...600x450.jpg",
  "page": 1,
  "scrapedAt": "2026-07-08T..."
}

Examples

All for-sale items in Brooklyn under furniture:

{
  "city": "newyork",
  "category": "sss",
  "searchQuery": "furniture",
  "maxPages": 5,
  "fetchBodies": true
}

Cars+trucks in Chicago, no body fetches (faster):

{
  "city": "chicago",
  "category": "cta",
  "searchQuery": "",
  "maxPages": 3,
  "fetchBodies": false
}

Browse all sublets in Seattle (no query):

{
  "city": "seattle",
  "category": "sub",
  "searchQuery": "",
  "maxPages": 2,
  "fetchBodies": true
}

Limits / gotchas

• 60 requests/min per IP (Craigslist's published limit; the actor respects this by default).
• maxPages hard-capped at 25 (input schema enforces 1-25). At 120 results per page that's 3000 listings per run.
• bodyExcerpt is null when fetchBodies=false, or for items past the first 5 per page even when fetchBodies=true (capped to keep request count bounded).
• priceValue is null when Craigslist shows no price (e.g. "Please contact" listings). price retains the raw display string.
• About 6% of result pages have "similar listings" sidebar links at the end of the href list. These don't have a JSON-LD partner and are skipped — they never make it into the dataset.

License

Actor code: MIT (inherited from project template).

Craigslist data: per Craigslist's terms of use, personal/non-commercial use is permitted. Commercial use cases (reselling listings, large-scale scraping for commercial datasets) should review CL's ToS and the specific city's robots.txt.

Changelog

• 0.1.0 — initial release. City + category + search query, JSON-LD + DOM-order pairing, optional body fetches, hard cap 25 pages.