Summary: SAP said on May 4, 2026 that it had signed a definitive agreement to acquire Prior Labs, a German startup only eighteen months old at the time, and that it plans to put more than €1 billion (about $1.16 billion) into the company across four years, shaping it into a frontier AI lab focused on the structured data that global business runs on. Prior Labs works on tabular foundation models (TFMs), AI that predicts directly from tables and databases — the very format in which enterprise information resides — and its open-source TabPFN has been downloaded more than 3 million times.
SAP said on May 4, 2026 that it had signed a definitive agreement to acquire Prior Labs, a German startup only eighteen months old at the time, and that it plans to put more than €1 billion (about $1.16 billion) into the company across four years, shaping it into a frontier AI lab focused on the structured data that global business runs on. The purchase price stays undisclosed; regulatory approval is still required, with closing expected in Q2 or Q3 2026.
The wager pointedly leaves out large language models. Prior Labs works on tabular foundation models (TFMs), AI that predicts directly from tables and databases — the very format in which enterprise information resides — and its open-source TabPFN has been downloaded more than 3 million times. In same-day coverage, TechCrunch also brought the story’s second half to light: SAP’s API policy bans OpenClaw and other non-endorsed agents, allowing only its own Joule and Nvidia’s NemoClaw.
The Terms: Founders Cash Out, and €1 Billion Follows Over Four Years
No price appears in the official materials. Pathfounders, however, has reported after checking with sources that the transaction is “almost all cash” and that the founders have already received “well over half a billion dollars” up front. Measured against the company’s funding record — a roughly $9.3 million pre-seed in February 2025 led by Balderton Capital — this is an enormous payoff earned on a very short clock. James Wise, partner at Balderton, described it as “one of Germany’s biggest ever venture outcomes.”
Inside SAP, Prior Labs will continue as an independent unit so that its research pace survives; turning its work into products will run through SAP AI Core, SAP Business Data Cloud, and the Joule agentic layer. SAP adds that it remains “fully committed” to the open-source strategy and developer ecosystem behind TabPFN.
A Freiburg Company Eighteen Months Old, and Its TabPFN
Three people started the company: Frank Hutter, who holds the CEO title, together with Noah Hollmann and Sauraj Gambhir. Freiburg is home to the headquarters, with additional offices in Berlin and New York, and employees arrived from Google, Apple, Amazon, Microsoft, G-Research, Jane Street, Goldman Sachs, and CERN. The scientific advisory board is set to add Turing Award winner Yann LeCun along with Bernhard Schoelkopf, director of the Max Planck Institute for Intelligent Systems.
The technical pedigree is likewise strong. Nature published the TabPFN series of papers, and the models have posted state-of-the-art results across hundreds of independent studies; the newest, TabPFN-2.6, holds first place on the benchmark TabArena, reaching instantly — in a single model — the accuracy a four-hour AutoML pipeline attains. Its hallmark is in-context learning: it can predict with no training whatsoever. Ahead lie scale-ups to millions of rows, real-time inference, and new data modalities.
The Case for Tables over LLMs
SAP CTO Philipp Herzig made the argument plainly: “Early on, SAP recognized that the greatest untapped opportunity in enterprise AI wasn’t large language models.” The company first tested that conviction with SAP-RPT-1, its own relational pretrained transformer, and is now reinforcing it through acquisition. Capabilities on the plan include a conversational interface that answers natural-language questions and what-if scenarios; GDPR-compliant predictions for scenarios such as payment delays, supplier risk, upsell, and churn; and agentic systems that bring tables, language, and images together while reasoning toward causation rather than correlation.
Interest in this category is climbing. Neuralk-AI, a competitor, has completed a raise as well (though smaller than Prior Labs’), and Fundamental exited stealth in February 2026 with $255 million in Series A funding. Where SAP holds the advantage is the data itself: the world’s business tables overwhelmingly reside inside SAP systems.
Agents: NemoClaw Clears the Bar, OpenClaw Does Not
Agents form the story’s other half. Under its API policy, SAP “prohibits” AI agents unless they fit “SAP-endorsed architectures” — a policy The Information noticed first. Two options have been cleared for use: the company’s own Joule Agents, still in beta, and NemoClaw, Nvidia’s enterprise-facing, security-centered OpenClaw deployment, which qualified because Joule supports the Nvidia Agent Toolkit announced in March.
Salesforce has taken the opposite route: under its new Headless 360 architecture, enterprises pick their own agents, and OpenClaw is among the choices. To the question of “who may operate APIs on the user’s behalf,” two enterprise-software giants now answer in opposite ways — something developers working on enterprise integrations should keep an eye on.
The Message for Europe’s AI Story
Hutter said joining SAP “gives us the resources, data environment and customer reach” to carry the category to its full potential, and the stated aim is to build a “globally-leading frontier AI lab for structured data — in Europe, in the open.” Where Europe’s AI story is usually narrated as either model research or buyouts by American giants, a combination of one of Germany’s biggest venture outcomes and €1 billion in long-term investment is real backing for applied frontier work that stays in Europe. The stakes for SAP are just as visible: with the stock squeezed by 2026’s “SaaSpocalypse,” CFO Dominik Asam told CNBC in January that what matters is how quickly SAP can adopt these technologies and defend its scale advantage. The shares edged up after the news.