One URL connects Claude or ChatGPT to Google Patents, Google Scholar and PubMed. Real counts and real records — not blocked pages, CAPTCHAs, or confident guesses.
https://www.bizgeniusapp.com/mcp/search.php
3 databases · 6 read-only tools · no API keys · 2-minute setup
Results are live — Google Patents, Google Scholar and PubMed, served by the same BizGenius pipeline your assistant gets once connected.
When Google blocks a crawler it doesn’t send an error — it sends a normal-looking page with nothing on it. An assistant scraping directly can’t tell “blocked” from “no results”, so it reports zero. Confidently.
How much has been published on hyperscanning?
fetch scholar.google.com → 200 OK 0 results parsedI couldn’t find any published work on hyperscanning. It may be a very new or niche topic.
How much has been published on hyperscanning?
count_google_scholar count_pubmedAbout 9,320 on Google Scholar. PubMed, which is exact, has 799 — the gap is preprints, conference papers and work outside biomedicine. By method, in PubMed:
Real figures, measured through these tools. PubMed counts are exact; the Scholar total is Google’s own estimate and shifts between identical requests — which the tools tell your assistant, every time.
A count and a search for every source. Counting is separate on purpose: most questions that sound like “find me papers” are really “how many are there?”
By keyword, inventor or assignee. De-duplicated across pages — Google repeats patents between pages, and naive tools count them twice.
Estimated totalsThe broadest net there is: preprints, theses, conference papers, books, and every field beyond biomedicine.
Estimated totalsBiomedical literature with full query syntax — field tags, booleans, date ranges. The only source here whose counts are precise.
Exact countsSign in with the BizGenius account you already have. No API key to create, copy or rotate.
BizGenius Patents/Scholar Search
https://www.bizgeniusapp.com/mcp/search.php
In ChatGPT, custom connectors need a Plus, Pro, Business or Enterprise plan.
One credit is one search request. Nothing expires mid-question, and you can see exactly where you stand at any time.
Need more, an invoice, or a seat for your whole team? Talk to us about Custom.
No. Every tool is read-only and marked as such. It searches public databases and returns results — there is no tool that writes, edits or deletes, and it has no access to your BizGenius citation library. That is a separate connector.
Because Google's own "About N results" figure is an estimate, and it changes between identical requests. We report it as approximate rather than dressing it up. When you need a precise number and the topic is biomedical, PubMed gives you an exact one for the same single credit.
On the free trial, searches stop and we tell you. On a paid plan, extra credits are billed automatically at your plan's rate — and we warn you before you get there and again when overage begins, so it is never a surprise on the invoice.
We record which tool was called, the query, result counts and timing — that is how usage is metered and how we diagnose failures. We never see your conversation with your assistant, only the search parameters. Full detail in the privacy policy.
Remove the connector in Claude or ChatGPT and access stops immediately. Or email us and we will revoke it from our side.
100 credits free. No card, no API key, about two minutes.
https://www.bizgeniusapp.com/mcp/search.php