Spanish startup Multiverse Computing raises €189 million to scale technology that compresses LLMs by up to 95%

San Sebastian-based Multiverse Computing, a global innovator in quantum-inspired AI model compression, today announces a €189 million investment round to further develop their CompactifAI – a compression tech capable of reducing LLMs by up to 95%.

The Series B will be led by Bullhound Capital with the support of world-class investors such as HP Tech Ventures, SETT, Forgepoint Capital International, CDP Venture Capital, Santander Climate VC, Quantonation, Toshiba and Capital Riesgo de Euskadi – Grupo SPRI.

“The prevailing wisdom is that shrinking LLMs comes at a cost. Multiverse is changing that,” said Enrique Lizaso Olmos, Founder and CEO of Multiverse Computing. “What started as a breakthrough in model compression quickly proved transformative – unlocking new efficiencies in AI deployment and earning rapid adoption for its ability to radically reduce the hardware requirements for running AI models. With a unique syndicate of expert and strategic global investors on board and Bullhound Capital as lead investor, we can now further advance our laser-focused delivery of compressed AI models that offer outstanding performance with minimal infrastructure.”

Founded in 2019, Multiverse Computing innovates quantum-inspired AI model compression. The company’s expertise in quantum software and AI led to the development of CompactifAI, an AI model compressor. CompactifAI compresses LLMs by up to 95% with only 2-3% precision loss. CompactifAI models reportedly reduce computing requirements and unleash new use cases for AI across industries.

With over 160 patents and 100 customers globally, including Iberdrola, Bosch, and the Bank of Canada, Multiverse Computing has raised near €216 million to date.

Per Roman, Co-founder & Managing Partner, Bullhound Capital, said: “Multiverse’s CompactifAI introduces material changes to AI processing that address the global need for greater efficiency in AI, and their ingenuity is accelerating European sovereignty. Román Orús has convinced us that he and his team of engineers are developing truly world-class solutions in this highly complex and compute intensive field. Enrique Lizaso is the perfect CEO for rapidly expanding the business in a global race for AI dominance.”

According to Multiverse Computing, LLMs typically run on specialised, cloud-based infrastructure that drives up data centre costs. Traditional compression techniques – quantisation and pruning – aim to address these challenges, but their resulting models significantly underperform original LLMs.

With the development of CompactifAI, Multiverse aims to address this. CompactifAI models are highly-compressed versions of leading open source LLMs that retain original accuracy, are 4x-12x faster and yield a 50%-80% reduction in inference costs. These compressed, affordable, energy-efficient models can run on the cloud, on private data centres or – in the case of ultra compressed LLMs – directly on devices such as PCs, phones, cars, drones and even Raspberry Pi.

Tuan Tran, President of Technology and Innovation, HP Inc., commented: “At HP, we are dedicated to leading the future of work by providing solutions that drive business growth and enhance professional fulfillment. Our investment in Multiverse Computing supports this ambition. By making AI applications more accessible at the edge, Multiverse’s innovative approach has the potential to bring AI benefits of enhanced performance, personalisation, privacy and cost efficiency to life for companies of any size.”

CompactifAI was created using Tensor Networks, a quantum-inspired approach to simplifying neural networks. Tensor Networks is a specialised field of study spearheaded by Román Orús, Co-founder and Chief Scientific Officer at Multiverse.

“For the first time in history, we are able to profile the inner workings of a neural network to eliminate billions of spurious correlations to truly optimise all sorts of AI models,” said Orús.

Compressed versions of top Llama, DeepSeek and Mistral models are available now, with additional models coming soon.

Damien Henault, Managing Director, Forgepoint Capital International, added: “The Multiverse team has solved a deeply complex problem with sweeping implications. The company is well-positioned to be a foundational layer of the AI infrastructure stack. Multiverse represents a quantum leap for the global deployment and application of AI models, enabling smarter, cheaper and greener AI. This is only just the beginning of a massive market opportunity.”