
Introducing Harrison.rad.1, the latest frontier in radiology-specific foundational models.
With increasing challenges from rising imaging volumes, workforce shortages and growing complexity of medical images, specialised capabilities from foundation models have the potential to accelerate development of tools to scale global healthcare. Harrison.rad.1 is a specialised multimodal LLM for radiology, it is not a medical device but rather an enabling technology that could support innovation across the radiology workflow. Launched in 2024, we are actively exploring its applications and welcome collaboration with partners to research this emerging field.
Harrison.rad.1 outperforms other frontier models by ~2x in the FRCR 2B Rapids exam.

Harrison.rad.1 excels in the same radiology exams taken by human radiologists, as well as in benchmarks against other foundational models.
The Fellowship of the Royal College of Radiologists (FRCR) 2B Rapids exam is considered one of the leading and toughest certifications for radiologists. Only 40-59% of human radiologists pass on their first attempt. Radiologists who re-attempt the exam within a year of passing score an average of 50.88 out of 60 (84.8%)*.
Harrison.rad.1 scored 51.4 out of 60 (85.67%). Other competing models, including OpenAI’s GPT-4o, Microsoft’s LLaVA-Med, Anthropic’s Claude 3.5 Sonnet and Google’s Gemini 1.5 Pro, mostly scored below 30**, which is statistically no better than random guessing.***
**Data on file
***As measured by mock examinations, actual FRCR 2B Rapids questions are not publicly available
FRCR Diagnostic Score
<Accelerating validation and responsible introduction of generative AI in healthcare.
Unmatched performance
- Excels in benchmarks against other foundational models, as well as in standard radiology examinations taken by human radiologists.
Revolutionary capabilities
- Harrison.rad.1 presents a wealth of potential opportunities in global healthcare, such as improving clinical excellence and quality, and providing non-medical use cases too, such as accelerating the development of AI products.
A specialised multimodal LLM for radiology
General purpose Large Language Models (LLMs) are powerful, but their broad, generic focus renders them less suited to critical applications where accuracy is of paramount importance. A highly specialised and nuanced function like healthcare requires a specialised model. This is why Harrison.rad.1 was born.
Harrison.rad.1 has several emergent capabilities:
What’s Harrison.rad.1’s big advantage?
Exclusive Dataset
We have proprietary access to extensive medical imaging data that is representative and diverse, enabling superior model training and accuracy.
Medical Expert Annotation
Our data is annotated at scale by medical specialists, ensuring high-quality training signals that enhance model reliability and clinical relevance.
Specialised Model Architecture
Our model is specifically designed for healthcare data, fine-tuned for clinical usefulness and factual correctness.
Want to get under the hood of Harrison.rad.1? Contact us.

CEO and Co-Founder
Harrison.ai
Harrison.rad.1 is a significant technological leap towards our end goal of creating dramatically more capacity in radiology. We are making our model available to select collaborators to help accelerate research into validation methods as a precursor to the responsible integration of HR1 into clinical or diagnostic practice.