AI stock trading startup EquiLibre raises funding at a $500M+ valuation

Prague-based artificial intelligence developer EquiLibre Technologies Inc. has raised funding at a valuation exceeding $500 million.

The company stated late Tuesday that the Series A round was led by Creandum, a Stockholm-based startup fund. The size of the raise was not disclosed. However, Creandum Vice President Cameron Sellers told TechCrunch that the firm’s investment was the largest it has ever contributed to a funding round.

The investor roster also includes Turning Prize laureate Richard Sutton, one of the investors of modern reinforcement learning. Sutton sits on the company’s advisory board alongside several other prominent computer scientists.

EquiLibre develops AI agents optimized for financial trading. The company launched its agents in 2025 with an initial focus on cryptocurrencies and later expanded its focus to traditional stock markets such as the Nasdaq. Today, its agents complete billions of dollars worth of trade per month. The company says it hasn’t yet closed a month with a loss.

Its software is based on technologies that its founders previously implemented in an AI model called DeepStack. Martin Schmid, Matej Moravcik and Rudolf Kadlec developed the algorithm in 2017 while pursuing their computer science doctorates.

DeepStack is optimized to play Texas hold ’em. The poker variant is a major focus of AI research because it’s an imperfect-information game, which means that players lack key information such as the strength of opponents’ hands. That uncertainty makes Texas hold ’em more difficult for AI models than games such as chess.

Information gaps are also a major factor in financial trading. That’s one of the reasons EquiLibre carried over key technologies from DeepStack to its AI agents. The company’s founders placed particular emphasis on a training method called self-play that they used to develop DeepStack. Self-play enables an AI model to learn how to play a game by playing millions of matches against itself.

Researchers have also applied the technology to other use cases besides financial trading and poker. Google DeepMind, for example, used self-play to develop a neural network called AlphaProof that is optimized to solve math problems. In 2024, it became the first AI to win an International Mathematical Olympiad medal.

“The question is no longer whether this approach works. It’s how big it can get,” said EquiLibre co-founder and Chief Executive Martin Schmid. “We’ve proven the technology in the world’s biggest and most liquid markets.”

EquiLibre will use most of the proceeds from its Series A round to grow its computing cluster. The company expects the new infrastructure to boost its trading agents’ profitability. EquiLibre will use the rest of the capital to hire more deep learning researchers and engineers.

Image: Unsplash