Prior Labs

Prior Labs builds multimodal tabular foundation models, starting with TabPFN, which make predictions on structured data without per-task tuning or ML pipelines. The product is offered as an open-source Python package, a REST API, an MCP server, and private cloud or on-prem deployments.

The company describes its customers as teams running prediction workloads on structured data across finance, healthcare, industrials, energy and tech, and lists case studies with banks, hospitals and manufacturers. Revenue comes from a free tier, usage-based Pro plans and commercial licences for Private Cloud or on-prem production use, priced by contact with sales.

The company is registered in Germany, lists offices in Berlin, Freiburg and New York, and also lists San Francisco among its locations. Its open job listings cover engineering, research, go-to-market and operations roles in those cities, including SAP partnership roles.

Product: TabPFN, a pre-trained tabular foundation model available via open-source Python package, REST API, MCP, and private cloud or on-prem deployment.