SAP is buying Freiburg-based Prior Labs, one of the leading startups in tabular foundation models, in what appears to be a move to push enterprise AI beyond chatbots and into the structured data that runs companies. Terms were not disclosed. However, well-place sources told Pathfounders that this is “almost all cash”.
SAP says Prior Labs will continue to operate independently, while SAP says it will invest more than €1 billion over four years to scale it into a frontier AI lab for structured business data. This is suggests this is an acqui-hire style deal.
The deal is expected to close in Q2 or Q3 2026, subject to regulatory approval.
The company was founded by Frank Hutter, Noah Hollmann and Sauraj Gambhir, and has drawn researchers from Google, Apple, Amazon, Microsoft, G-Research, Jane Street, Goldman Sachs and CERN. Yann LeCun and Bernhard Schölkopf are set to join its scientific advisory board.
Prior Labs had previously raised €9M in funding over one Pre-Seed round raised in February 2025 with Balderton Capital. Guy Podjarny and Peter Sarlin are the most recent investors. Others in the round were include, Atlantic Labs, Mandeep Singh, XTX Ventures, Thomas Wolf, Galion.exe, Steve Anavi, Christopher Lynch, Edward Grefenstette, Robin Rombach, Ashutosh Kulkarni, and the Hector Foundation.
Balderton partner James Wise commented on X confirming the acquisition: “SAP have just announced their intent to acquire Prior Labs - in one of Germany’s biggest ever venture outcomes. It’s been a delight working with them over the last 18 months, from founding to exit. Far more importantly, this is a huge investment in what will be Germany’s most important frontier AI lab, bringing together Prior Lab’s exceptional team and technology with SAP’s global expertise and resources.”
In a relatively unnoticed Tweet from January, Wise had also commented that Prior Labs’s latest model for tabular data had “outperformed SOTA benchmarks for time series data, capabilities that even surprised the founding team.”
Clearly, SAP thinks the next frontier in enterprise AI is in the data that exists in spreadsheets and tables.
Large language models are powerful at text, but less reliable when it comes to the rows, columns, numbers and statistical patterns inside enterprise systems. Prior Labs builds tabular foundation models, or so-called TFMs, designed to predict outcomes from structured business data: late payments, supplier risk, churn, upsell potential, demand shifts and more.
In a statement, SAP CTO Philipp Herzig said "Early on, SAP recognised that the greatest untapped opportunity in enterprise AI wasn't large language models; it was AI built for the structured data that runs the world's businesses," SAP CTO Philipp Herzig said. "We built SAP-RPT-1 to prove that conviction for enterprise data. Prior Labs has built a leading TFM on public benchmarks and built one of the leading research teams in this category. Combining their frontier model work with enterprise data and customer reach is how we intend to lead this category globally."
Prior Labs’ says its open-source TabPFN tool has had more than 3 million downloads. Its latest model, TabPFN-2.6, is described as the top-performing model on TabArena, matching the accuracy of a four-hour automated machine learning pipeline almost instantly.
SAP appears to be making a bet on both the product bet and datas sovereignty in Germany and Europe, becasue it wants Prior Labs’ models wired into SAP AI Core, SAP Business Data Cloud and Joule, its agentic AI layer.
The idea is that business users will be able to ask natural-language questions, run “what-if” scenarios and get predictions from company data without training a bespoke model or relying on a team of data scientists.
