How to Scrape Walmart in 2026 (API & Python)

The fastest way to scrape Walmart in 2026 is to call a structured API that returns normalized JSON — search results, product detail, and reviews — instead of reverse-engineering the internal endpoints behind walmart.com. Walmart has no self-serve public API for outside developers: its old Open API is closed to new signups and its current Marketplace API only exists for approved sellers managing their own listings. This guide covers all three approaches, what each returns, where DIY breaks, and the legal reality up front.

Why scrape Walmart data?

• Price monitoring — track how price and price_text move on specific items over time.
• Competitor tracking — compare assortment, brand mix, and seller_name (Walmart.com vs. third-party marketplace sellers) across a category.
• Market research — see how search ranks products for a query and how availability shifts.
• Review and sentiment research — analyze rating, review_count, and per-star breakdowns to gauge how a product is actually received.
• AI pipelines — feed normalized product and review JSON into an agent or RAG pipeline without a page-parsing layer.

Is it legal to scrape Walmart?

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

Walmart's storefront renders from internal APIs rather than static HTML, so a DIY scraper has to reverse-engineer the same calls the site's own frontend makes:

import requests

# Walmart's storefront calls internal, undocumented endpoints —
# not a licensed public API, and robots.txt disallows /api/ and /search
resp = requests.get(
    "https://www.walmart.com/ip/Apple-AirPods-Pro-2/5689919121",
    headers={"User-Agent": "Mozilla/5.0 (compatible; research-bot/1.0)"},
)

It demos and then breaks:

• The Terms of Use are explicit. Walmart names "robot, spider... scrape, data mine" tools directly and requires prior written consent for any automated gathering of site content.
• robots.txt disallows the paths a scraper needs. /api/, /search, and /typeahead/ are all off-limits to crawlers by Walmart's own directive.
• Real, PerimeterX-class anti-bot at the edge. Walmart runs bot-detection and challenge pages in front of its storefront; naive requests calls get blocked or served CAPTCHA-style interstitials fast.
• No official, documented product API. The data behind every page comes from internal endpoints not published or licensed for outside use, and their shape shifts with redesigns without notice.
• Everything fans out per item. Full detail and reviews each live behind a separate call keyed by item_id — one page load never gets you a complete product record.

Option 2: No-code / ready-made tools

Marketplace scraper actors exist for one-off Walmart pulls, but they inherit the same Terms-of-Use exposure and anti-bot fragility as DIY, and still leave you stitching together search, detail, and review calls yourself to assemble a full product record.

Option 3: A structured Walmart API

For a repeatable, structured workflow, a Walmart scraping API returns normalized JSON with no page parsing or edge-defense upkeep. Search the catalog:

curl "https://api.crawlora.net/api/v1/walmart/search?q=wireless+earbuds" \
  -H "x-api-key: $CRAWLORA_API_KEY"

Then resolve an item_id and pull detail and reviews in Python:

import requests

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

hits = requests.get(f"{base}/search", headers=h, params={"q": "wireless earbuds", "page": 1}).json()["data"]["items"]
item_id = hits[0]["item_id"]

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

A search response is normalized JSON (real fields — check the docs):

{
  "code": 200,
  "msg": "OK",
  "data": {
    "query": "wireless earbuds",
    "page": 1,
    "total_results": 1200,
    "items": [
      {
        "item_id": "6543706564",
        "title": "Bose QuietComfort Earbuds, White Smoke",
        "brand": "Bose",
        "canonical_url": "https://www.walmart.com/ip/6543706564",
        "price": 179,
        "price_text": "$179.00",
        "currency": "USD",
        "availability": "IN_STOCK",
        "seller_name": "Walmart.com",
        "rating": 4.4,
        "review_count": 1203
      }
    ]
  }
}

page and sort are optional query params for /walmart/search, so you can paginate a full result set and re-sort by price or rating without changing the query. Product detail nests price, availability, and attributes by item_id:

{
  "data": {
    "item_id": "5689919121",
    "title": "Apple AirPods Pro 2",
    "brand": "Apple",
    "price": 169,
    "price_text": "$169.00",
    "availability": "IN_STOCK",
    "seller_name": "Walmart.com",
    "rating": 4.6,
    "review_count": 17823,
    "attributes": { "Brand": "Apple" }
  }
}

And reviews return an aggregate breakdown plus individual entries:

{
  "data": {
    "average_rating": 4.3,
    "total_review_count": 2452,
    "recommended_percent": 87,
    "rating_counts": { "five_star": 1698, "four_star": 273, "three_star": 150, "two_star": 116, "one_star": 215 },
    "reviews": [
      { "rating": 5, "text": "Love these earbuds.", "author": "MusicFan", "submission_date": "11/30/2025", "review_id": "406499690" }
    ]
  }
}

Store one row per item_id and re-run on a schedule to track price and availability changes.

What you can collect

Public catalog data per endpoint: paginated, sortable search results (items, total_results, page); product detail (title, brand, price, price_text, availability, seller_name, rating, review_count, description, attributes); and review data (average_rating, total_review_count, recommended_percent, per-star rating_counts, and individual reviews with rating, text, author, submission_date). Public catalog data only.

Limitations

• No self-serve public API. The Walmart Open API is closed to new registrations and the Marketplace API is gated to approved sellers and solution providers — nothing here is a licensed path to bulk catalog access.
• Real anti-bot at the edge. Walmart runs PerimeterX-class bot detection and challenge pages in front of the storefront; naive requests get blocked fast.
• The Terms of Use are explicit. Walmart names robots, spiders, and scraping/data-mining tools directly, and separately restricts using site content for AI/ML training.
• Per-item_id fan-out. Full detail and reviews each need a separate call keyed by item_id — there's no single call that returns a complete product record.
• No store- or zip-level pricing. Search and product responses return one national price and availability state, not per-store variance.
• Public data only. This is catalog data Walmart already shows publicly — never a way to reach account, order, or post-login information, or to bulk-download the catalog wholesale.

Where this gets used

• Price and deal tracking — watch price and price_text move on tracked items and flag markdowns.
• Competitive benchmarking — compare assortment and brand mix against other big-box and marketplace retailers.
• Review research — track rating, review_count, and star-distribution trends by product.
• Sourcing and market research — spot which sellers (seller_name) are winning placement on high-volume search terms.

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 search and product endpoints in the Playground, check the schema in the API docs, and review pricing. For the same "no self-serve API, gated developer program" pattern on another retail giant, see how to scrape Amazon product data; for the big-box-retail version of this same problem, see how to scrape Target or how to scrape Costco. See also is web scraping legal.

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