Pricing

from $6.50 / 1,000 product page scanneds

Try for free

Go to Apify Store

Product Spec Table Extractor

Try for free

Extract product specification tables, attribute lists, dimensions, compatibility data, materials, warnings, and download or manual links from product pages.

Pricing

from $6.50 / 1,000 product page scanneds

Rating

0.0

(0)

Developer

junipr

Actor stats

Bookmarked

Total users

Monthly active users

17 days ago

Last modified

What it does

Extracts product attributes from HTML specification tables, definition lists, and Product JSON-LD additionalProperty records, with normalization and provenance.

The actor emits one row per extracted product specification, or a no-spec diagnostic. Every dataset row uses the declared dataset schema, includes source provenance and a scan timestamp, and is written only after every applicable pay-per-event charge is accepted.

Typical uses

Extract product specification tables
Build catalog attribute exports
Find missing product specs
Collect product manual links
Normalize product dimensions and units

Input

Use public HTTP(S) URLs or deterministic htmlInputs. URL fetching rejects localhost, loopback, link-local, and private-network destinations. allowedDomains can further constrain live fetches.

startUrls
sitemapUrls
htmlInputs
allowedDomains
maxPages
maxDepth
includeEvidence
includeRawSnapshots
requestDelayMs
timeoutMs
userAgentMode
productUrls
attributeNameMap
includeDownloads
normalizeUnits
maxSpecRowsPerProduct
tableSelectors
maxProductPages
maxSpecRowsPerPage
maxTablesPerPage
maxChargeUsd
includeReport

The safety caps named in maxProductPages, maxSpecRowsPerPage, maxTablesPerPage, timeoutMs are enforced in addition to maxChargeUsd. A value of 0 for maxChargeUsd is a hard zero budget: the actor exits before the first paid operation.

Minimal deterministic example

{
  "htmlInputs": [
    {
      "sourceUrl": "https://shop.example/public-page",
      "html": "<html><head><title>Public page</title></head><body><h1>Public evidence</h1></body></html>"
    }
  ],
  "includeEvidence": true,
  "includeReport": true,
  "maxChargeUsd": 15
}

The prefilled input remains deliberately small so Apify automated quality runs finish quickly. For production use, replace the example source with public pages you are authorized to access.

Dataset output

sourceUrl
sourceType
pageTitle
recordType
recordName
issueCode
severity
evidenceSnippet
recommendation
canonicalUrl
httpStatus
scannedAt
productUrl
productName
attributeName
attributeValue
attributeGroup
normalizedValue
unit
downloadUrl
manualUrl
extractionConfidence
extractionSource
specIssue

Reports are also written to key-value storage:

product-spec-table-extractor-summary-report.md
product-spec-table-extractor-records.csv
product-spec-table-extractor-issues.json
product-spec-table-extractor-evidence-samples.json
product-spec-table-extractor-run-metadata.json

When includeRawSnapshots is true, bounded raw HTML snapshots are stored separately. Raw snapshots are off by default.

Pay-per-event pricing

Platform-usage pass-through is off. Prices are the locked fixed-inclusive values for this actor.

Event	Price	Meaning
`actor-start`	$0.01	Actor Start
`product-page-scanned`	$0.0065	Product Page Scanned
`spec-row-extracted`	$0.00295	Spec Row Extracted
`spec-table-issue-emitted`	$0.00295	Spec Table Issue Emitted
`audit-report-generated`	$0.08	Audit Report Generated

Charging rules:

actor-start must be accepted before fetching or rendering begins.
Page/target events are deduplicated per source.
Record and finding events fire only when the corresponding output exists.
Report output is written only after the report event is accepted.
Partial charges and Apify charge-limit responses stop the run before additional paid output.
Store discounts are tracked as a separate Console configuration requirement; this source does not claim they are already applied.

Evidence and failure behavior

Each row identifies its extraction source and confidence.
Duplicate attributes are marked rather than silently discarded.
Unit normalization is conservative and preserves the raw value.

A fetch failure, browser failure, blocked page, missing record, or incomplete match produces a bounded diagnostic row when possible. It never invents a successful extraction. Evidence can be omitted with includeEvidence: false.

Local verification

pnpm --filter @junipr/product-spec-table-extractor build
pnpm --filter @junipr/product-spec-table-extractor test
pnpm --filter @junipr/product-spec-table-extractor smoke

The tests cover strong, weak, unavailable, comparison, and cap fixtures; schema equality; pricing metadata; accepted/partial/limited charges; zero budget; and report generation.

Limitations

Results reflect the public source at the scan time and may change later.
Site markup changes can require selector/parser maintenance.
The actor does not bypass authentication, CAPTCHAs, robots restrictions, or access controls.
Legal, tax, pricing, SEO, merchandising, and marketplace decisions remain the user's responsibility.

Amazon Product Scraper

mrdoe/amazon-product-scraper

Scrape Amazon product listings and extract structured product data including titles, prices, ratings, reviews, availability, product URLs, images, and other product details. Search by keyword and collect product data for price monitoring, product research, market analysis, and e-commerce automation.

MrDoe

Amazon Product Scraper🛍️

webscrap18/amazon-product-scraper

🛒 Scrape Amazon Product Data with Ease Easily extract comprehensive product information from Amazon.com, including search results, product pages, and category listings. Get access to 💰 pricing details, ⭐ customer reviews & ratings, 📦 product availability, and 🏷️ key product specifications.

WebScrap

5.0

MCP 2026 Compatibility & Migration Auditor

obliging_persimmon_cki/mcp-2026-compatibility-migration-auditor

Bounded, explainable audits of public MCP endpoint compatibility against a pinned official specification.

Dung Huynh

HTML Table Extractor

benthepythondev/html-table-extractor

Extract structured rows from HTML tables on any web page.

Ben

Table Extractor — Scrape HTML Tables from Any URL to JSON/CSV

eliai/webpage-tables-extractor

Table extractor for webpages: pass a URL and get every HTML data table back as structured rows — JSON via the API or dataset, CSV/Excel via one-click export. Layout tables skipped automatically. For analysts, developers and AI agents. $0.02 per page, failures free.

Broke to Built

Wikipedia Table Extractor

lafuan/wikipedia-data

Extract structured data from Wikipedia tables as clean JSON. Supports all languages, multiple tables and header detection. Ideal for AI training data and research.

Muhammad Naufal

Target Product Scraper

parseforge/target-scraper

Scrapes product data from Target.com category and product pages. Returns each product as a flat row with price, rating, and description when details are enabled.

ParseForge

5.0

Product Review Summary Extractor

junipr/product-review-summary-extractor

Extract public product reviews, ratings, dates, pros/cons snippets, verified-buyer signals, and theme summaries from product or review pages

junipr

Amazon Product Data Extractor

extractlab/extractlab-amazon-product-extractor

Extract complete Amazon.com product data (price, images, videos, ratings, specs, shipping, variants) by ASIN, product URL, or keyword search.

Extract Lab

Naver Shopping Product Detail Scraper (네이버쇼핑) - Specs

zen-studio/naver-product-detail-scraper

Scrape complete Naver Shopping product data by link or ID: price, discount, stock, specification tables, every seller's price, six months of price history, images, category path and the seller's own product record. Works with price-comparison and store links alike. Export to JSON, CSV or Excel.