Fractal's Vaidya 2.0 becomes the first AI to score 50.1 on OpenAI's HealthBench (hard), outperformin...
The AMW Read
The launch of a domain-specific model outperforming general-purpose frontier models on specialized benchmarks validates the verticalized foundation model pattern and aligns with India's sovereign AI policy/mission.
NoveltySignificance
Healthcare & Bio · Player MapGeopolitics
Fractal's Vaidya 2.0 becomes the first AI to score 50.1 on OpenAI's HealthBench (hard), outperforming GPT-5 and Gemini Pro 3. 🏥 This specialized healthcare model proves domain-specific AI can now surpass general-purpose models in clinical reasoning. With India's INR 10,300 Cr AI mission behind it, this launch marks a shift toward verticalized foundation models built for population-scale healthcare in emerging markets. 🌏
