Spanish artificial intelligence startup, Multiverse Computing, has secured €189 million ($217 million) in funding from Bullhound Capital, HP Inc., Forgepoint Capital, and Toshiba to advance its innovative AI compression technology. The firm announced on Thursday that the breakthrough allows large language models (LLMs) to be shrunk by up to 95% without sacrificing performance, cutting operational costs by as much as 80%.
Multiverse’s approach uniquely blends principles from quantum physics and machine learning to simulate quantum systems—without the need for quantum computers. This has positioned the company as a pioneer in efficient AI infrastructure, particularly in an era where model size and computational demand are key challenges for widespread adoption.
With this investment, Multiverse becomes Spain’s largest AI startup and joins the ranks of top European AI innovators like Mistral, Aleph Alpha, Synthesia, Poolside, and Owkin. The company has already released compressed versions of open-source LLMs such as Meta’s Llama, China’s DeepSeek, and France’s Mistral, with more in the pipeline.
CEO Enrique Lizaso Olmos said the company is focusing on compressing the most widely used models in corporate environments, notably the Llama family. Their AI compression tools are also available on Amazon Web Services’ AI marketplace, offering scalable access to businesses looking to cut costs and boost performance in AI applications.