News Summary

Spanish startup Multiverse Computing has successfully raised $215 million in a Series B funding round, aimed at advancing its innovative AI model compression technology named ‘CompactifAI’. This quantum-computing inspired technology can reduce the size of large language models (LLMs) by up to 95% without sacrificing performance. The company’s ‘slim’ models are significantly faster and more cost-effective, offering a 50-80% reduction in inference costs. These models are accessible via Amazon Web Services or can be licensed for on-premise use, and are versatile enough to run on a variety of devices including PCs and Raspberry Pi. Multiverse Computing, co-founded by CTO Román Orús and CEO Enrique Lizaso Olmos, has already secured 160 patents and serves 100 clients globally, including the Bank of Canada. The funding round, led by Bullhound Capital, aims to further the deployment of this groundbreaking technology.