Multiverse Raises $215M for AI Model Compression

Multiverse Computing, a Spanish quantum-inspired AI startup, has raised approximately $215 million in Series B funding led by Bullhound Capital, joined by investors such as HP Tech Ventures, Forgepoint Capital, and Toshiba. The company's proprietary CompactifAI technology can compress large language models (LLMs) by up to 95% while maintaining comparable accuracy and reducing inference costs by up to 80%. This enables LLMs to run more efficiently on a wide range of devices, including PCs, smartphones, and even Raspberry Pi, promoting broader and more sustainable AI adoption. Multiverse now targets the compression of popular open-source LLMs such as Meta’s Llama, Mistral, and DeepSeek, with CompactifAI available on AWS and for on-premises use. The funding round raised Multiverse’s valuation to over $500 million—a fivefold increase since its last round—making it the largest AI startup in Spain. The company aims to lower AI deployment costs and support sustainability in AI computing.