Prior to this, the company secured $215 million in June 2025.
Bloomberg News previously covered the fundraising.
Multiverse seeks to lower the cost and improve the efficiency of AI deployment for businesses by compressing large language models to reduce their energy and computational demands. The Spanish government had previously announced its intention to participate in this funding round.
Get the market news that matters in a five-minute read with Market Briefs, our free daily newsletter
Multiverse is among a rising group of startups that pledge to reduce the recurring costs of running AI models - what are referred to as inference costs. Among the startups that raised notable funding rounds this year are OpenRouter (a model marketplace), Runpod (a developer platform), and Baseten (an inference infrastructure platform), according to PitchBook data.
Multiverse's efficiency technology, dubbed CompactifAI, relies on tensor networks. Tensor networks, commonly employed in quantum computing, work by selectively simplifying data to facilitate large-scale calculations. When applied to AI, this approach accelerates model performance and improves efficiency while maintaining high accuracy, according to the company.
Multiverse Computing was founded by a team with expertise in quantum physics and machine learning. Its tensor-network approach, originally developed for quantum simulations, has found a novel application in compressing neural networks. This cross-domain innovation has attracted both corporate and sovereign investors seeking to reduce the environmental impact of large AI models.
The drive to reduce inference costs is accelerating as enterprises scale AI deployments. By leveraging tensor networks, Multiverse Computing aims to make large models more economical for businesses, potentially expanding the market for AI applications. The company's approach aligns with growing demand for sustainable AI solutions that lower energy consumption without sacrificing performance.
Founded in 2019, Multiverse Computing has positioned itself at the intersection of quantum computing and artificial intelligence. The company's use of tensor networks - a method originally developed for simulating quantum systems - has drawn interest from both government and corporate investors eager to reduce the energy footprint of large AI models. With this latest capital injection, Multiverse plans to expand product development, scale its engineering team, and accelerate global sales efforts.
Other backers include Banco Santander SA, Orange Ventures, the Basque government, and the Qatar Development Bank, along with additional participants. Advisers to Multiverse included Banco Santander SA and JPMorgan Chase & Co.
The startup's focus on inference cost reduction places it in a fast-growing niche. As enterprises adopt generative AI at scale, the operational expense of running models repeatedly has become a critical bottleneck. Multiverse's tensor-network approach offers a potential solution by compressing models without significant accuracy loss, which could make AI deployment more accessible to mid-sized companies and industries with tight margins.
The company's ability to attract both corporate venture arms and sovereign wealth funds underscores the strategic importance of energy-efficient AI infrastructure. With the new funding, Multiverse is well-positioned to compete against larger players in the inference optimization space while maintaining its roots in quantum-inspired techniques.
Join Market Briefs, our free daily newsletter, for a quick daily rundown of the markets
