Bay Marketplace Insights API Alternative (2026): Sold Data Without eBay's Approval

Looking for an eBay Marketplace Insights API alternative? You're in good company. Marketplace Insights is the only official eBay API that returns sold items, and it's a limited release: you apply, you explain your business case, and most of us never hear back or get turned down.

Meanwhile the product you're building still needs sold prices. This post covers what eBay actually offers, why each official route dead-ends, and a working alternative you can call today.

What eBay offers for sold data in 2026

eBay API	Sold prices?	Can you use it?
Marketplace Insights	Yes, about 90 days	Limited release, approved partners only
Browse API	No, active listings only	Yes
Finding API (findCompletedItems)	Used to	Decommissioned February 2025
Trading / Sell APIs	Only your own sales	Yes, for your own account
Terapeak (Product Research)	Yes	Browser dashboard only, no API

So if your app needs other people's sold prices, eBay leaves you with nothing you can call.

The alternative: one GET request

CompSniper is an eBay sold listings API that needs no eBay approval, no developer account and no OAuth dance. Sign up, copy a key, call it.

curl -H "Authorization: Bearer $COMPSNIPER_API_KEY" \
  "https://api.compsniper.com/v1/scrape?keyword=nintendo+switch+oled&count=240&ebaySite=ebay.com"

You get up to 240 sold listings with the sold price, sold date, condition, shipping and seller, plus a summary with the median, quartiles, min, max and average shipping.

How it compares with what Marketplace Insights would give you

	Marketplace Insights	CompSniper
Access	Apply and wait for approval	Free key in a minute
Sold history	About 90 days	Recent sales, plus up to 3 years with /v1/research/sold
Results	Up to 10,000 per search, paged	Up to 240 per request, more with Max Mode
Price summary	Build it yourself	Median and range in every response
Wrong matches	Raw keyword results	Optional AI cleanup removes accessories, parts and wrong models
Best Offer sales	Not documented	Real accepted price with bestOfferPrice=true
eBay sites	Selected marketplaces	8: US, UK, Germany, France, Italy, Spain, Canada, Australia
Other marketplaces	No	Poshmark, Mercari, TCGplayer, Grailed and more on the same key

On accuracy: in a blind 40-search benchmark, 5.4% of the sales CompSniper returned were the wrong product, against 25.7% to 31.6% for other eBay sold data APIs tested. That matters more than it sounds. Wrong matches (a case, a charger, a different storage size) are exactly what drag a median away from the real price.

Python: recent sold prices

import os
import requests

r = requests.get(
    "https://api.compsniper.com/v1/scrape",
    headers={"Authorization": f"Bearer {os.environ['COMPSNIPER_API_KEY']}"},
    params={"keyword": "nintendo switch oled", "count": 240, "ebaySite": "ebay.com"},
    timeout=90,
)
r.raise_for_status()
data = r.json()

s = data["summary"]
print(f"{s['count']} sales, median {s['median']} {s['currency']}, middle half {s['p25']} to {s['p75']}")
for item in data["items"][:5]:
    print(item["endedAt"], item["soldPrice"], item["condition"], item["title"][:60])

Or pip install compsniper for a ready-made client with retries.

Going past 90 days

This is where you'd hit Marketplace Insights' ceiling even if you were approved. /v1/research/sold returns sold history between any two dates, up to 3 years back:

r = requests.get(
    "https://api.compsniper.com/v1/research/sold",
    headers={"Authorization": f"Bearer {os.environ['COMPSNIPER_API_KEY']}"},
    params={
        "keyword": "nintendo switch oled",
        "startDate": "2023-11-05",
        "endDate": "2026-11-04",
        "sort": "lastSoldDescending",
        "limit": 50,
    },
    timeout=60,
)
r.raise_for_status()
for row in r.json()["results"]:
    print(row["lastSoldDate"], row["soldUnits"], row["averageSoldPrice"]["amount"], row["title"][:60])

Page through with page and hasNextPage. Every plan, free included, gets the full 3 years.

JavaScript (Node 18+)

const params = new URLSearchParams({ keyword: "nintendo switch oled", count: "240", ebaySite: "ebay.co.uk" });
const res = await fetch(`https://api.compsniper.com/v1/scrape?${params}`, {
  headers: { Authorization: `Bearer ${process.env.COMPSNIPER_API_KEY}` },
});
if (!res.ok) throw new Error(`API error ${res.status}`);
const data = await res.json();
console.log(data.summary.median, data.summary.currency); // prices in GBP on ebay.co.uk

Each eBay site returns prices in its own currency: GBP, EUR, CAD, AUD or USD.

Migrating code that expected Marketplace Insights

If you designed around Marketplace Insights' item_sales/search, the mapping is straightforward:

• q becomes keyword
• X-EBAY-C-MARKETPLACE-ID: EBAY_GB becomes ebaySite=ebay.co.uk (the eBay IDs like EBAY_GB are accepted too)
• lastSoldPrice becomes soldPrice, and lastSoldDate becomes endedAt
• the OAuth client-credentials flow becomes one Authorization: Bearer header

Keep your own interface in front of it (get_sold_listings(keyword, site)), and if eBay ever approves you, swapping the backend is a one-file change.

Try it

Free key at compsniper.com: 100 requests a month, no credit card, and 3-year history included. Docs: API reference and 3-year sold history. Code: ebay-sold-listings-api and ebay-sold-history-api on GitHub.

Did you ever get a reply on a Marketplace Insights application, or did it just sit in the queue like most?