PARTLY

Partly true: Ultimate Web Scraper really does plug into Claude as a custom MCP connector and pull a product catalog into a table, but 'unlimited data extraction from any website' is the vendor's ad copy and cloud scraping is metered per page

Facebook · Oct 2026

The claim'Claude users are living in 2078.' Go to Claude, click customize, add custom connector, open the Ultimate Web Scraper Chrome extension, copy its MCP URL, paste it into Claude. 'With this tool Claude can now scrape any website, so you just give it a link and it will scan the entire catalog and give you all the product data including the images and then of course you can export it.' Caption: 'Claude + Ultimate Web Scraper = unlimited data extraction from any website.'

The demo is real and the screen recording is a faithful walkthrough. Ultimate Web Scraper is a Chrome extension from Panda Tech with a free local-extraction tier and a cloud platform behind it; its site advertises a Claude connection where you ask in natural language, the scrape runs in the cloud, and a structured table comes back. The reel shows exactly that: the extension's MCP dialog (mcp.ultimatewebscraper.com), Claude's Add custom connector form, a prompt pointing at a Breitling collection page, a 66-row product table with names, descriptions, prices and images, and an Export menu with CSV, Excel, JSON and Google Sheets. So Claude plus this connector can pull a product catalog into a table. Two things are oversold. 'Unlimited' and 'any website' are the vendor's words, and the vendor's own site admits LinkedIn is not supported by any of its tools; sites with aggressive bot protection, login walls or terms that forbid scraping are not a given, and the cloud runs that make the Claude flow work are charged per page from prepaid credits. The free plan covers local extraction in the browser, not the AI-agent connection shown in the reel, which lives on the paid plans. Also note that Claude is not doing the scraping; it is relaying a request to the vendor's cloud and formatting the result, so your Claude usage and the vendor's credits both get spent. This is first-party advertising presented as a discovery, which is fine as long as you read 'living in 2078' as an ad.

What holds up

Reel read frame by frame: Claude's sidebar with past scrape chats, Settings > Connectors > Add custom connector, the extension's 'Add Ultimate Web Scraper to Claude' dialog with the mcp.ultimatewebscraper.com URL, a prompt against a Breitling Navitimer collection URL, a 66-row cloud table with images and prices, and an Export menu (CSV, Excel, JSON, Google Sheets, Copy to Clipboard).
ultimatewebscraper.com read: Chrome, Edge and Brave extension by Panda Tech, Inc.; 'connects to Claude and other MCP-compatible AI applications'; free plan is unlimited local extraction with 'some power features limited'; paid plans add the cloud platform, AI agent connections, scheduling and integrations; cloud scraping is charged per page from prepaid credits with a cost estimate before each run.
The same site states 'LinkedIn isn't supported by any of our tools,' which already contradicts 'any website'; the connector depends on the vendor's cloud, not on Claude, so bot-protected or login-walled targets succeed or fail on the vendor's side.
The poster is the product's own Facebook page; the reel is a first-party ad with no price shown.

What doesn’t

'Unlimited data extraction from any website' is ad copy; the vendor's own FAQ names at least one unsupported site and meters cloud runs per page.
The Claude-connected flow shown runs on the paid cloud tier; the free tier is local extraction only.
Claude does not scrape anything itself; it calls the vendor's service, so you pay Claude usage and vendor credits for the same job.
Scraping a retailer's catalog can breach that site's terms; the reel never raises it.

The catch

It is a real, convenient way to turn a catalog page into a spreadsheet from inside Claude, and the export options are genuine. Budget for per-page cloud credits, expect some sites to refuse, and remember that a scraped catalog is someone else's content; use it for research and pricing checks, not for republishing.

How to actually do it

Install the extension, run a free local extraction on one catalog page first, and confirm the fields you need (price, SKU, image URL) come through before paying for anything.
If you need the Claude flow, add the connector in Claude Settings > Connectors, run one small collection, and check the credit cost estimate the vendor shows before the run.
Export to CSV or Sheets and keep the scrape for internal use; check the target site's terms before you build anything customer-facing on the data.

Ultimate Web Scraper's Claude connector works as demonstrated and exports real tables. The 'unlimited, any website' framing is the maker's own marketing, the AI-connected tier is paid and metered, and the reel is first-party advertising.

Confidence: High
Posted by: Ultimate Web Scraper (Panda Tech, Inc.), the product's own page; reel sent to Buddy Thu Sep 24, 2026, ~9:36 PM

See the original claim →

We test hype for free. We build the real thing for a living.

Thirty minutes, no pitch — and you'll leave with something useful either way.

Book a call with Todd or start with the free Business Checkup →

The Verdict Weekly

Three verdicts every Friday. Free forever, unsubscribe anytime, no spam — that would be ironic.