Prior Labs, an 18-month-old Berlin AI startup, just sold to SAP for over $1.1 billion. Here's how tabular foundation models are reshaping enterprise AI and what it means for European tech.
### The Fastest Billion-Dollar Exit in European AI?
Berlin-based Prior Labs, a frontier AI lab specializing in tabular data, just pulled off something remarkable. Eighteen months after being founded, the company has been acquired by SAP in a deal backed by over $1.1 billion in investment. That's right, over a billion dollars for a startup that was still a research project just a year and a half ago.
And here's the kicker: Prior Labs will keep operating under its own brand, with its own leadership and research agenda. They'll continue publishing their work and making their models openly available. All of this follows their $9.7 million pre-Seed round back in 2025.
"Eighteen months ago, Prior Labs was a research project," says Frank Hutter, co-founder and CEO. "Today we're beginning our next chapter as an AI lab with the resources to tackle problems we simply couldn't before."
### Why Tabular Foundation Models Matter
So what exactly does Prior Labs do that's worth over a billion dollars? They've pioneered something called tabular foundation models, or TFMs. Think of it as AI built specifically for the kind of data that powers most businesses: spreadsheets, databases, and structured records.
Most of the AI hype has been around large language models (LLMs) that process text. But Prior Labs took a different bet. They focused on the structured data that runs the world's businesses. Their model, TabPFN, can handle prediction tasks like forecasting payment delays, customer churn, supplier risk, and demand directly from enterprise data.
"Early on, SAP recognized 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," adds Philipp Herzig, CTO of SAP.
### Real-World Impact Already Happening
This isn't just theoretical. Prior Labs' technology is already:
- Helping prevent train failures with Hitachi
- Improving financial forecasting with TD Bank
- Applied across hundreds of research projects, from pancreatic cancer diagnosis to wildfire prediction and next-generation battery materials
The acquisition lets Prior Labs deploy these models across SAP's massive enterprise customer base, which includes some of the largest companies on the planet.
### The Bigger Picture for European AI
SAP's investment in Prior Labs is part of a much larger wave of funding flowing into European AI infrastructure. EU-Startups' 2026 coverage shows approximately $2.1 billion across comparable companies, led by Nscale's $1.85 billion AI-compute round and Verda's $109 million financing.
Germany alone has produced several direct competitors, including Berlin's SPREAD, Cognee, and Qorelo, plus Stuttgart's Blockbrain and Munich's Interloom. Including SAP's commitment to Prior Labs, the total activity referenced represents more than $3.2 billion.
### What This Means for Enterprise AI
This acquisition is a signal that enterprise AI is entering a new phase. Instead of training separate models for every dataset, companies can now use a single pre-trained foundation model. That's a massive efficiency gain.
For Prior Labs, the deal provides access to enterprise data environments and long-term investment that would have been impossible for an 18-month-old startup. They can now pursue multi-year frontier research programs across enterprise AI, scientific discovery, causality, relational data, and agentic systems.
"Taking tabular foundation models to the next level requires better data environments, deployment surfaces, and long-term research investment," Hutter says. "SAP is uniquely positioned to provide all of these."
### Looking Ahead
Prior Labs was founded in 2024 by Frank Hutter, Noah Hollmann, and Sauraj Gambhir. With SAP's backing, they're now aiming for even more ambitious projects, including what they call "moonshots" in medical data and material sciences.
It's a story that shows how quickly things can move in AI. One moment you're a research project. The next, you're backed by one of the world's largest software companies with over a billion dollars to spend. For European startups, that's a powerful example of what's possible.