How to Scrape Whatnot in 2026 (API & Python)

The fastest way to collect Whatnot data in 2026 is a structured API that returns live-show and category listings as normalized JSON. Whatnot is a live-commerce and auction platform — sellers run video streams where buyers bid or buy in real time — which is a genuinely different genre from a static product catalog like eBay or StockX. This guide covers what's actually scriptable: which shows are live, what categories exist, and a show's product/listing metadata. It does not cover real-time in-stream bidding activity (the live bid ticks inside a stream), which isn't exposed by any public interface.

Why scrape Whatnot data?

• Live-commerce market research — track how many shows are running in a category at a given time, and which sellers are active.
• Category and seller trend tracking — watch which categories (trading cards, sneakers, women's fashion) are growing in show volume.
• Collectibles market monitoring — trading cards, graded slabs, and sneaker resale are the platform's core categories; tracking show titles and tags surfaces what's trending.
• AI pipelines and agents — feed normalized show and category data into a market-monitoring or research agent instead of screen-scraping a video-first UI.

Is it legal to scrape Whatnot?

Option 1: DIY in Python (and why it breaks)

A naive scraper hits the browse page and tries to pull show cards out of rendered markup:

import requests
from bs4 import BeautifulSoup

resp = requests.get(
    "https://www.whatnot.com/category/trading_card_games",
    headers={"User-Agent": "Mozilla/5.0 (compatible; research-bot/1.0)"},
)
soup = BeautifulSoup(resp.text, "html.parser")
# Live show cards, thumbnails, and status render from client-side state —
# you end up parsing JSON out of a script tag, not stable HTML

It demos once, then breaks:

• The ToS is explicit. Whatnot names scraping, spidering, and crawling directly as a prohibited use of the App — this isn't ambiguous.
• The page is a live app, not a document. Show status (PLAYING vs. ended), viewer counts, and product state update via websocket/client state, not static markup — a UI change silently breaks a CSS-selector scraper.
• There's no public read API to fall back on. The official Seller API is a closed Developer Preview for sellers managing their own inventory — it isn't a substitute for browsing the public catalog.
• Anti-bot defenses. Like most high-traffic commerce apps, Whatnot fronts its site with bot detection; scripted requests without a real browser session get rate-limited or blocked quickly.

Option 2: No-code / ready-made tools

Browser extensions and point-and-click scrapers can export a single browse page's show list, but tracking a fast-moving live-show platform means re-pulling category listings on a schedule to catch shows starting and ending — a job for a stable endpoint, not a one-off manual export.

Option 3: A structured Whatnot API

For a repeatable workflow, a Whatnot scraping API returns normalized JSON for the category list, live shows by category, and a show's product detail — no page parsing, no closed developer program to apply to. Start with the category list:

curl "https://api.crawlora.net/api/v1/whatnot/categories" \
  -H "x-api-key: $CRAWLORA_API_KEY"

{
  "code": 200,
  "msg": "OK",
  "data": {
    "categories": [
      { "id": "Q2F0ZWdvcnlOb2RlOjE0OQ==", "name": "Trading Card Games", "slug": "trading_card_games" },
      { "id": "Q2F0ZWdvcnlOb2RlOjM=", "name": "Sports Cards", "slug": "sports_cards" },
      { "id": "Q2F0ZWdvcnlOb2RlOjY1Ng==", "name": "Women's Fashion", "slug": "womens_fashion" }
    ]
  }
}

Use a category slug to browse its live shows, then pull a specific show's products by id:

import requests

h = {"x-api-key": "YOUR_API_KEY"}
base = "https://api.crawlora.net/api/v1/whatnot"

categories = requests.get(f"{base}/categories", headers=h).json()["data"]["categories"]

browse = requests.get(
    f"{base}/browse", headers=h, params={"category": "trading_card_games"}
).json()["data"]
show_id = browse["shows"][0]["id"]

show = requests.get(f"{base}/live/{show_id}", headers=h).json()["data"]

browse returns the shows currently associated with a category:

{
  "code": 200,
  "msg": "OK",
  "data": {
    "category": "trading_card_games",
    "shows": [
      {
        "id": "6ddb7fd2-43bb-44e0-8ee8-6656e82fa26a",
        "url": "https://www.whatnot.com/live/6ddb7fd2-43bb-44e0-8ee8-6656e82fa26a",
        "title": "$100k Late Night Slab Show",
        "seller_username": "legacy_auction_house",
        "status": "PLAYING",
        "start_time_ms": 1785996739803,
        "thumbnail_url": "https://images.whatnot.com/...",
        "tags": ["Sudden Death", "Graded Cards", "Pokémon"]
      }
    ]
  }
}

And live/{id} returns that show's product listings (check the docs for the full field list):

{
  "code": 200,
  "msg": "OK",
  "data": {
    "id": "6ddb7fd2-43bb-44e0-8ee8-6656e82fa26a",
    "url": "https://www.whatnot.com/live/6ddb7fd2-43bb-44e0-8ee8-6656e82fa26a",
    "products": [
      {
        "id": "TGlzdGluZ05vZGU6MjEzMTA5NzI1NA==",
        "title": "slab",
        "description": "show live",
        "price_cents": 1000,
        "currency": "USD",
        "status": "running",
        "transaction_type": "AUCTION",
        "quantity": 219,
        "seller": { "username": "legacy_auction_house", "rating": 4.9, "review_count": 128914 }
      }
    ]
  }
}

What you can collect

• Categories: name and slug for every top-level category Whatnot organizes shows under (trading cards, sports cards, fashion, and more)
• Browse-by-category: currently associated shows — ID, URL, title, seller username, status, start time, thumbnail, and descriptive tags
• Live show detail: a show's product/listing entries — title, description, price, currency, status, transaction type (auction vs. fixed price), quantity, and basic seller stats (username, rating, review count)

This is catalog and show-level metadata — a snapshot of what's listed and running, not a stream of individual bid events.

Limitations

• The API surface is narrow. This is a 3-endpoint group — browse, categories, and live show detail — there's no search, no seller-profile endpoint, and no historical archive of past shows.
• No in-stream bidding data. Bid-by-bid activity inside a live auction (who bid what, when, the running high bid tick-by-tick) isn't exposed by any public interface, including this one — live/{id} returns listing state, not a bid log.
• Show state is a snapshot. status, quantity, and price fields reflect the moment you pull them; a fast-moving auction show can change seconds later.
• No official public read API to compare against. Whatnot's own Seller API is a closed preview for sellers managing inventory, so there's no first-party alternative for general market research.
• Public data only. This covers public browse and show pages — not account data, DMs, order history, or a way to place a bid on a seller's behalf.

Where this gets used

• Collectibles and resale market research — track which categories and tags (graded cards, specific franchises) show up most often across live shows.
• Live-commerce trend tracking — monitor show volume and seller activity by category over time.
• Feed enrichment for research and monitoring agents — normalized show/category JSON instead of parsing a video-first web app.

Sources

Start collecting

Try it first, free: run any public URL through the Free Web Scraper, or check whether a site blocks bots with the Anti-Bot Checker — no signup.

Test the categories, browse, and live-show endpoints in the Playground, check the schema in the API docs, and review pricing. Whatnot's live-auction model sits next to two other collectibles and resale marketplaces in this series: how to scrape StockX for order-book resale pricing, and how to scrape eBay for fixed-price and traditional-auction listings. See also is web scraping legal.

Part of our how-to-scrape guide series — every platform we cover, in one index.