Nvidia teams up with Abridge to build AI model for healthcare: report

Nvidia teams up with Abridge to build AI model for healthcare: report
Vatsala Gaur
11 Jun 2026, 16:01 PM

powered by

Invezz
Nvidia (NVDA)

Buy NVDA. The partnership with Abridge is a concrete win in healthcare AI workflows (clinical notes + decision support) using Nvidia’s Nemotron open models. It reinforces Nvidia as the default AI infrastructure provider for vertical, data-trained models—exactly where budgets are shifting from generic demos to deployed products.

Key Risk: Hospitals and health systems slow adoption or demand cheaper alternatives, cutting Nvidia’s ability to monetize healthcare AI infrastructure.

Abridge (ABR)

Buy Abridge. Nvidia’s model integration plus Abridge’s hospital-grade ambient listening creates a defensible product loop: better clinical transcription → better training data → better notes/workflows. If Abridge becomes the leading “clinical conversation” layer, it can expand from documentation into real-time clinical support and capture more of the workflow spend.

Key Risk: Abridge’s model performance doesn’t beat incumbents in real hospital use (accuracy, safety, and clinician trust), limiting deployments despite the Nvidia partnership.

  • Nvidia and Abridge are building a healthcare-specific AI model.
  • The partnership highlights growing competition in healthcare AI.
  • Abridge will use de-identified clinical data to train the model.

Nvidia is expanding its presence in healthcare artificial intelligence through a new partnership with medical technology startup Abridge, as competition intensifies among technology companies seeking to apply AI tools across the healthcare industry.

According to a report by The Wall Street Journal, Nvidia and Abridge are developing an AI model specifically designed for clinical conversations between doctors and patients.

The model will be deployed exclusively within Abridge's platform and is intended to improve medical documentation, clinical decision support, and other healthcare workflows.

Nvidia is already an investor in the Pittsburgh-based startup, which has emerged as one of the leading providers of AI-powered note-taking tools for healthcare professionals.

Focus on clinical intelligence

The new model will be built using Nvidia's Nemotron family of open AI models and then customized using Abridge's healthcare expertise and clinical data.

Kimberly Powell, Nvidia's vice president of healthcare, said the partnership reflects a broader effort to adapt AI models for specialized industries.

"There's an opportunity now to take these models and adapt them with this clinical intelligence at a much earlier stage of model development," Powell said.

Unlike general-purpose AI models, the healthcare-focused system is being trained to understand the nuances of doctor-patient interactions, medical terminology and clinical decision-making.

Powell said the collaboration also demonstrates how Nvidia's open models can be tailored for sectors such as drug discovery, medical devices, and digital health.

"Nvidia has been known to be the AI company that works with every other AI company," Powell added.

Healthcare becomes the next AI battleground

The announcement comes amid a broader race among technology firms to build healthcare-focused AI products.

Last week, Microsoft unveiled a collaboration with Mayo Clinic aimed at creating a healthcare-specific AI model trained on clinical data.

OpenAI and Anthropic have also introduced healthcare offerings targeted at both consumers and medical professionals.

The growing interest reflects the size of the healthcare market and the potential for AI to reduce administrative burdens, improve productivity, and assist with clinical decision-making.

For Abridge, the partnership offers access to Nvidia's AI infrastructure while allowing the company to incorporate years of clinical expertise into the model.

Abridge brings medical data and expertise

Abridge plans to use de-identified clinical data gathered through its platform to further train and customize the model.

"Generic models are powerful, but clinical intelligence it still has to be trained, it has to be shaped, and it has to be evaluated against real-world conditions. That's a lot of our early experiments and work with the Nemotron team," said Dr. Shiv Rao, Abridge's co-founder and chief executive.

Founded in 2018, Abridge specializes in ambient-listening technology that records conversations between doctors and patients and automatically generates clinical notes.

The company raised $300 million last year in a funding round that valued it at $5.3 billion, underscoring investor enthusiasm for healthcare AI applications.

The new model is expected to become one of several AI systems powering Abridge's platform when it launches later this year.

According to Davis Liang, Abridge's director of applied science, the company previously developed its own speech-recognition model because existing third-party systems were not optimized for hospital environments and medical terminology.

Cost efficiency is another factor driving the partnership.

Liang said smaller, purpose-built open models are often less expensive than proprietary alternatives and can be deployed on Abridge's own infrastructure.

Hospitals see potential beyond documentation

Healthcare providers are already exploring how the technology could evolve beyond note-taking.

Dr. Joon Lee, chief executive of Emory Healthcare, said the health system has rolled out Abridge's platform to more than 3,000 physicians.

Lee believes increasingly sophisticated AI tools could eventually provide real-time support during patient consultations.

"The doctor-patient conversation is at the center of so much decision-making and information exchange," Lee said.

"I'm excited about the possibility of intelligence coming more and more in real-time."

As healthcare becomes one of the most competitive frontiers for artificial intelligence, Nvidia's latest partnership signals that the battle for medical AI leadership is only beginning.