Multiverse Computing raises $215M for tech that could radically slim AI costs

Spanish startup Multiverse Computing secured €189 million in Series B funding to advance its "CompactifAI" technology, which compresses large language models (LLMs) by up to 95% without compromising performance. The company offers compressed versions of open-source LLMs, claiming 4x-12x faster speeds and a 50%-80% reduction in inference costs, with applications ranging from cloud services on AWS to edge deployment on devices like PCs and Raspberry Pis. Led by Bullhound Capital, the funding round included participation from HP Tech Ventures, Santander Climate VC, and Toshiba, bringing Multiverse's total funding to approximately $250 million.

Analysis

Spanish startup Multiverse Computing has secured a significant €189 million (approximately $215 million) Series B funding round, underscoring robust investor confidence in its "CompactifAI" technology. This quantum-computing inspired compression technology claims to reduce the size of Large Language Models (LLMs) by up to 95% without impacting model performance, a critical advancement in the AI field. Multiverse offers compressed "slim" versions of popular open-source LLMs such as Llama and Mistral variants, with DeepSeek R1 support forthcoming, stating these models are 4x-12x faster and achieve a 50%-80% reduction in inference costs; for example, its Lama 4 Scout Slim costs 10 cents per million tokens on AWS compared to the original's 14 cents. These efficiencies enable deployment not only on cloud platforms like Amazon Web Services and on-premises but also on edge devices including PCs and Raspberry Pis, significantly broadening the accessibility of advanced AI. The company, co-founded by CTO Román Orús, a pioneer in tensor networks, and CEO Enrique Lizaso Olmos, with a background in mathematics and banking, has attracted a strong syndicate of investors led by Bullhound Capital—known for backing firms like Spotify—and including HP Tech Ventures, Santander Climate VC, and Toshiba. With 160 patents, 100 global customers such as Bosch and the Bank of Canada, and approximately $250 million raised to date, Multiverse is positioning itself as a key player in making AI models more accessible and cost-effective, reflecting the "strongly positive" sentiment surrounding this development in AI technology and innovation.

Key Decisions for Investors

• Investors should recognize the growing importance of AI model optimization, as demonstrated by Multiverse Computing's substantial Series B funding, and assess opportunities in companies developing or leveraging such efficiency-enhancing technologies.
• Consider the potential downstream impact on the AI value chain, including reduced operational costs for enterprises deploying LLMs and increased viability of edge AI applications, which could influence investment strategies in related hardware, software, and cloud service sectors.
• Private equity and venture capital investors should note the strong backing from established firms like Bullhound Capital and strategic corporates like HP Tech Ventures and Toshiba, indicating significant perceived value and growth potential in the LLM compression market, a key area within the broader Artificial Intelligence theme.