Vet-AI's automated triage app offers instant advice to pet owners 24/7, early detection of health issues and advice as to whether they need to speak to a vet.
For the blinded test, which was conducted in April 2025, three UK veterinary surgeons reviewed the responses from Vet-AI, Gemini 2.0 and ChatGPT 4.0 to 48 simulated enquiries from owners.
The vets evaluated each on clinical accuracy, triage effectiveness and the subjective criteria of factuality, integrity, safety, empathy and readability.
Vet-AI’s Large Language Model (LLM) achieved an accuracy score of 81%, surpassing Gemini (69%) and ChatGPT (50%).
Vet-AI’s model also stopped triage at the correct point in 81% of cases according to the vets, compared with Gemini’s 75% and ChatGPT’s 56%.
Vet-AI’s tool also outperformed the leading LLMs on subjective measures, delivering more empathetic, thorough and factual responses to pet owners.
Samantha Webster, Chief Veterinary Officer at Vet-AI said: “These latest results are a significant milestone in the development of automated triage.
"We’re proving we can provide trustworthy online pet care by combining clinical and technical expertise.
"This tool can support the wider industry too, engaging more pet owners in the first step of vet care, ensuring their pets get the care they need - whether in practice or remotely.
"We know that an increasing percentage of the UK pet population is struggling to access veterinary care.
"Whilst AI will never replace a vet, the use of innovative technology can break down the barrier to access, leading to more pets getting the care they need, when they need it."
“With this level of clinical accuracy and technical innovation, we're helping to bring great pet care to more people. As well as helping to create more demand for vet services by making the triage of pets as accurate as possible,” said Paul Hallett, Vet-AI founder and exec chair.
https://www.vet-ai.com/automated-triage
PS: Whilst you're here, take a moment to see our latest job opportunities for vets.