Prior Labs builds foundation models that understand spreadsheets and databases, a domain that has remained largely untouched as foundation models revolutionised text and images. The company's flagship model, TabPFN, achieved state-of-the-art performance on tabular machine learning and was published in Nature. With over 2.5 million downloads and 5,500 GitHub stars, TabPFN has established significant technical credibility among practitioners.
The company applies this research across scientific discovery, medical research, financial modelling, and business intelligence. TabPFN's performance on tabular data - which underpins much of the world's critical research and decision-making - addresses a genuine gap in how foundation models serve real-world analytics and discovery workflows.
Prior Labs assembles engineers and researchers from leading institutions and scales rapidly under backing from Balderton Capital, XTX Ventures, and the Hector Foundation. The company is advised by Yann LeCun, the Turing Award-winning computer scientist. Current work extends toward agentic AI systems capable of multi-modal fusion - integrating tables with language and images - and incorporating domain knowledge to serve specialised research and business needs.
