Zorgm Pro achieves 98.54% score on NEET PG benchmark using source-grounded clinical AI: LaennecAI team

Image for Zorgm Pro achieves 98.54% score on NEET PG benchmark using source-grounded clinical AI: LaennecAI team

A medical AI just sat one of India's toughest exams.

It didn't just pass.

It scored 98.54%.

152 out of 154 questions. Correct.

One wrong. One unanswered.

That's it.

The AI is called Zorgm Pro, built by a quiet medical AI company called LaennecAI โ€” with teams scattered across London, Cardiff, Birmingham, and Kochi.

And the exam? NEET PG โ€” the gateway every Indian doctor sweats through to enter postgraduate specialist training.


๐Ÿง  Why this number is genuinely wild

NEET PG isn't a multiple-choice trivia quiz.

It's the wall 2,30,114 doctors crawled through in 2025โ€ฆ and only ~1.28 lakh cleared.

Years of medical school. Sleepless nights. Coffee IVs.

And an AI just casually outscored almost every human who took it.


โšก But here's the actual plot twist

The score isn't the headline.

The way it got there is.

Most AI tools โ€” ChatGPT, Gemini, the usual suspects โ€” work like confident interns.

They answer fast. They sound sure.

And sometimesโ€ฆ they hallucinate.

In medicine, a hallucination isn't a cute bug.

It's a patient.

Zorgm Pro took a different route: Retrieval Augmented Generation (RAG) strapped to a curated library of:

  • ๐Ÿ“š Peer-reviewed journals
  • ๐Ÿฉบ Clinical guidelines
  • ๐Ÿ’Š Regulator-approved drug labels
  • ๐Ÿ‡ฎ๐Ÿ‡ณ National protocols like NACO India

Every answer comes with the receipt attached.


๐ŸŽฏ A real example from the test

One question hit a brutal topic: paediatric HIV vertical-transmission prophylaxis.

Zorgm Pro didn't guess.

It pulled India's PPTCT guidance, NIH Perinatal Guidelines, and FDA-approved references โ€” and cited them in the answer.

That's not a chatbot. That's a junior doctor with a library card.


๐Ÿ’ธ And the price tag?

Zero.

Free for verified physicians. No paywall. No tier upgrades.

๐Ÿ‘‰ In a country where elite medical coaching costs lakhs, a tool scoring 98.54% on NEET PG is being handed to doctors at no cost.

As LaennecAI's CTO Dr. Arathy Varghese put it:

"An answer is only as good as the evidence behind it. Ours comes with the source attached every time, so a doctor never has to take it on faith."


๐Ÿš€ The bigger signal

Clinical AI has been stuck in a trust crisis for two years.

Impressive demos. Scary errors. Hospitals hesitant.

Zorgm Pro's run suggests the unlock isn't a bigger model.

It's a smaller, source-grounded one that knows when to look things up.

The future of medical AI may not be the smartest doctor in the room.

It'll be the most honest one.

That's all for now!