Krux
February 21, 2026
Fractal's Vaidya 2.0 Beats GPT-5 on Healthcare Benchmark
Published: February 21, 2026 at 3:43 AM
Updated: February 21, 2026 at 3:43 AM
100-word summary
Fractal just launched Vaidya 2.0, a healthcare AI model that's crushing the competition. It scored 50.1 on OpenAI's notoriously tough HealthBench test, becoming the first model to cross 50 and outperforming both GPT-5 and Google's Gemini Pro 3. Think of it as an AI doctor's assistant that can triage emergencies, check symptoms, and guide patients through treatment. Unlike typical AI models that just recall medical facts, Vaidya 2.0 actually reasons through complex healthcare scenarios. Part of India's $1.2 billion AI push, it's designed specifically for healthcare systems in developing regions, bridging the gap between raw patient data and actual medical action.
What happened
Fractal just launched Vaidya 2.0, a healthcare AI model that's crushing the competition. It scored 50.1 on OpenAI's notoriously tough HealthBench test, becoming the first model to cross 50 and outperforming both GPT-5 and Google's Gemini Pro 3. Think of it as an AI doctor's assistant that can triage emergencies, check symptoms, and guide patients through treatment. Unlike typical AI models that just recall medical facts, Vaidya 2.0 actually reasons through complex healthcare scenarios.
Why it matters
Part of India's $1.2 billion AI push, it's designed specifically for healthcare systems in developing regions, bridging the gap between raw patient data and actual medical action.