eBay Pro Scraper — Browse API + Web Fallback

Extract eBay item data — prices, conditions, sellers, listing URLs — from the official eBay Browse API (pure HTTP, reliable, never blocked) with an optional HTML/web fallback for sold/completed listings.

⚠️ 2026 reality: eBay now requires a signed-in session for the sold/completed search view (LH_Sold=1) — anonymous clients (plain HTTP or a real headless Chrome through a proxy IP) are redirected to the sign-in wall or an Akamai "orch" verify-yourself captcha (splashui/captcha). There is no setting that disables this. This is exactly why the actor is session-first:

1. Fast primary path (recommended): provide a logged-in eBay session cookie (env EBAY_SESSION_COOKIE, or auto-login, or KV cache) → full LH_Sold=1 list is fetched with all 240 sold cards (price, shipping, sold date) in ~20-25s. Verified end-to-end.
2. Auto-login (ebayUsername/ebayPassword): mints a fresh session, verifies it against the sold SRP, caches it in the key-value store (3h) and reuses it next runs. Works reliably only when the sign-in origin IP is not risk-flagged (no captcha). From flagged residential/datacenter IPs eBay answers with the "orch" verify-yourself captcha before the form, and the login cannot complete — in that case use a cookie minted from your own browser or a trusted IP.
3. Sessionless fallback: if no session is available, the actor skips the dead anonymous-SOLD probe and recovers sold prices via public seller feedback profiles (FEEDBACK_COMP: real per-transaction prices + item IDs) and ended-item PDPs ("This listing sold on …").

The Browse API path uses the official API for active listings (pure HTTP, never blocked).

Data Source (source input)

source	Behavior
auto (default)	Uses Browse API if ebayClientId + ebayClientSecret are present, otherwise falls back to web
browse_api	Official eBay API only. Requires free keys (see below). Recommended.
web	HTML scraping incl. the sold-data fallback: active SRP → seller feedback profiles (real transaction prices) + ended-PDP enrichment. Runs automatically on 0 items.

Getting free Browse API keys (2 minutes)

1. Go to developer.ebay.com → sign in with any eBay account.
2. Create an application (choose "Production" keys when done).
3. Copy the Client ID and Client Secret into the actor input fields (ebayClientId, ebayClientSecret) or the env vars EBAY_CLIENT_ID / EBAY_CLIENT_SECRET.
4. Done — no billing. The Browse API has generous free call quotas for research volume.

Input

Field	Type	Default	Notes
keyword	string	—	Search keyword (e.g. rtx 4080)
startUrls	array	—	eBay URLs; _nkw= is used as the query in Browse API mode
count	integer	20	Max items to return
source	string	auto	auto / browse_api / web
ebayClientId	string	—	Browse API app key (secret input)
ebayClientSecret	string	—	Browse API app secret (secret input)
detailedSearch	boolean	false	Reserved for web mode details
testItemIds	array	—	Ended/active item IDs to probe on their PDP pages (sold price + end date)
recon	boolean	false	Run slow eBay JS-recon diagnostics (outputs WEB_RECON)
scoutBypass	boolean	false	Probe mobile + seller-scoped (_ssn=) sold-search bypasses; pushes real sold items as WEB_SOLD when a variant works
ebayUsername	string	—	eBay login username (secret). With ebayPassword, the actor logs in automatically (via real Chromium first, curl fallback), verifies a sold-unlocked session cookie and caches it in the key-value store (3h). Caveat: auto-login can only complete when the run's exit IP is NOT risk-flagged — otherwise eBay answers with an "orch" verify-yourself captcha and no session is minted. Most reliable option: set the env var EBAY_SESSION_COOKIE to a cookie harvested from your own logged-in browser (Actor → Settings → Environment variables); it is used automatically and persisted to the KV cache.
ebayPassword	string	—	eBay login password (secret); used with ebayUsername

With a valid session (auto-login, KV-cached, or env cookie) the actor pushes real sold/completed listings as WEB_SOLD (soldPrice, soldDateText, shipping, total) from the LH_Sold=1 search page — a fast primary comp source (~20-25s). Without one it degrades to seller-feedback comps (FEEDBACK_COMP).

Example Input

{
  "keyword": "rtx 3080",
  "count": 50,
  "source": "browse_api",
  "ebayClientId": "YourClientId",
  "ebayClientSecret": "YourSecret"
}

Output Fields

Field	Type	Description
itemId	string	Legacy item ID (Browse API) or URL-derived ID (web)
url	string	Item view URL
title	string	Listing title
soldPrice	string	Price (Browse API: current price; web: sold price)
soldCurrency	string	Currency code (USD, GBP, EUR…)
shippingPrice	string	Shipping cost (web mode only)
totalPrice	string	Price + shipping (web mode only)
condition	string	Item condition
sellerName	string	Seller username (Browse API)
itemEndDate	string	Listing end date (Browse API)
itemCreationDate	string	Listing creation date (Browse API)
scrapedAt	string	Extraction timestamp

FAQ

Does this return sold/completed items?

The Browse API path returns current/active listings (the official API has no sold-search since the Finding API was retired in Feb 2025). The web fallback (source: web, or auto with no keys) recovers sold prices with the seller-feedback technique: anonymous active search yields sellers, their public feedback profiles expose real per-transaction prices + item IDs, and ended-item PDPs confirm the sold price/end date. It runs automatically when the normal scrape returns nothing. Rows are tagged type: "FEEDBACK_COMP" (per-transaction sold comps) plus a PDP_FALLBACK_REPORT diagnostic row.

Why did you add a browser-free TLS impersonation tool (curl-impersonate)?

It is the community-standard way to make plain HTTP look like a real Chrome browser. Combined with residential proxies + warmup it succeeds on active SRP and feedback pages (the paths that matter for the sold-data fallback); sold/completed search still returns the sign-in wall.

Legal & Compliance Disclaimer

This actor is an independent tool and is not affiliated with, endorsed by, or sponsored by eBay. It accesses publicly available listing data via eBay's official API and/or public web pages. You are responsible for compliance with eBay's Terms of Service and applicable law.

SEO Keywords

ebay scraper, ebay browse api, ebay price tracker, ebay market data, ebay completed listings, ebay api, product pricing data, resale research, price analysis, price monitoring, ebay data extraction