Spanish Startup Multiverse Computing Targets 570 Million Dollar Round to Shrink AI Models Using Quantum Physics

Multiverse Computing, a Spanish startup applying mathematical techniques from quantum physics to the problem of AI efficiency, is raising up to 570 million dollars in a Series C funding round at a pre‑money valuation of 1.7 billion dollars, a five‑fold increase over its Series B valuation from just over a year earlier.

The round is co‑led by Forgepoint Capital International, BNP Paribas Solar Impulse Venture Fund and Bullhound Capital, with additional participation from Santander Alternative Investments, Tikehau Capital, HP, Orange Ventures, Scania Invest, Qatar Development Bank and several other strategic and sovereign investors. The round remains open to additional strategic backers, and once finalized is expected to bring Multiverse Computing's total funding raised to approximately 800 million dollars.

Compressing AI Models Using Physics, Not Just Engineering Tricks

At the center of Multiverse Computing's platform is CompactifAI, a compression technology built around tensor networks, a mathematical framework originally developed within quantum physics to describe complex, highly interconnected systems efficiently. Applied to large language models, the technique can reduce a model's size by 80 to 95 percent while preserving the vast majority of its original accuracy, according to the company, allowing compressed versions of popular open‑source models such as Llama and Mistral to run considerably faster and at meaningfully lower cost.

The technology was pioneered by Multiverse co‑founder and chief scientific officer Dr Román Orús, an Ikerbasque research professor at the Donostia International Physics Centre, working alongside chief executive Enrique Lizaso Olmos, a former deputy chief executive at Unnim Bank. That pairing of a physicist focused on the underlying mathematics with an executive experienced in scaling a financial institution has shaped Multiverse's approach to commercializing what began as fundamentally academic research into quantum‑inspired computation.

A Practical Answer to Rising AI Compute Costs

Multiverse's central argument is that the AI industry has, for years, accepted an unnecessary constraint: the assumption that powerful AI models require correspondingly expensive infrastructure to run. By compressing models to a fraction of their original size without materially compromising their output quality, CompactifAI allows organizations to deploy capable AI systems locally on smartphones, industrial equipment and other edge devices, including environments where cloud connectivity is unavailable, prohibitively expensive, or restricted under data sovereignty regulations.

That positioning has proven especially relevant to organizations operating in regulated or infrastructure‑constrained environments. Multiverse already counts more than 100 global customers, including energy utility Iberdrola, automotive supplier Bosch and the Bank of Canada, spanning sectors such as manufacturing, finance, energy, aerospace, cybersecurity, defense and healthcare, where the ability to run AI models locally, securely and without dependence on constant cloud connectivity carries particular value.

Revenue Growth That Outpaced the Company's Own Projections

Since closing its Series B round of 189 million euros in June 2025, Multiverse says its annualized revenue has increased more than tenfold, with first‑quarter 2026 sales growing 96 times year over year. That scale of revenue acceleration, while still measured off a relatively small initial base, has been a significant factor in justifying the company's steep valuation increase between funding rounds, particularly at a moment when investors across the AI sector have grown more attentive to demonstrated commercial traction rather than technology potential alone.

The new capital will fund continued research and development on Multiverse's proprietary compression algorithms, expansion of its library of pre‑compressed, efficient AI models, and strategic investment in what the company describes as sovereign AI gigafactory infrastructure, alongside the software stack needed to support broader enterprise deployment. Multiverse also plans to strengthen its regional presence across East Asia, Southeast Asia, the Middle East, Canada and the United States as part of the round.

Part of a Broader Race to Make AI Cheaper to Run

Multiverse's fundraise lands amid intensifying investor interest in AI efficiency and compression technology more broadly, a category that has drawn significant capital as the costs of running large‑scale AI models at inference time continue to climb across the industry. The scale of Multiverse's round, and the presence of sovereign and strategic investors including a national development bank, reflects how seriously governments and large enterprises are now treating AI compute efficiency as a matter of both cost control and national infrastructure independence, rather than a narrow technical optimization problem confined to individual companies.

With a five‑fold valuation increase, a broadening customer base across critical infrastructure sectors, and revenue growth that has significantly outpaced its prior funding cycle, Multiverse Computing's Series C positions the company as one of the more closely watched bets in the growing race to make AI not just more capable, but fundamentally cheaper and more portable to run.