Multiverse Computing, a Spanish startup, has raised $217 million in a round led by Bullhound Capital, HP Inc., Forgepoint Capital, and Toshiba. The goal of the investment is to scale its innovative compression technology for AI models. The company, which was started by Enrique Lizaso Olmos, says it can make big language models smaller. It will reduce operating costs by 80% and complete tasks by up to 95% without compromising accuracy.
Additionally, the company’s strategy integrates concepts from quantum computing, even though it makes use of conventional hardware. Thus, Multiverse is now the biggest AI company in Spain and one of the leading companies in open-source AI development in Europe.
Quantum Physics Helps Redefine Language Model Efficiency
Multiverse Computing has created a novel technique for AI model compression. It will make use of concepts taken from machine learning and quantum computing. This method does not require a quantum computer, in contrast to real quantum systems. Moreover, the compression system reduces the size of large language models without affecting their functionality. The process involves mimicking certain aspects of quantum psychology.
The company launches these compressed models on Amazon Web Services. Multiverse has already made versions of compressed language models, such as China’s DeepSeek, France’s Mistral, and Meta’s Llama, available. The goal is to optimize only the most used open-source AI models that are already in use across major corporations.
How Multiverse Is Transforming Open-Source AI Tools?
According to Multiverse’s CEO, Enrique Lizaso Olmos, the company is improving on existing AI models rather than developing new ones. “Most companies we speak to are using the Llama family of models,” he noted. Therefore, their method guarantees seamless integration for businesses that do not want to completely redesign their current systems.
Multiverse now has the same level of funding as other well-known European startups. Among the startups are Synthesia, Owkin, Poolside, and Aleph Alpha. In contrast to many of these companies, Multiverse does not develop full-scale models; instead, it concentrates only on compression technology. Their tools are especially helpful for companies that need open-source AI solutions that are lighter and faster. Additionally, these tools will shorten deployment times and lower server costs.
Future Outlook for Compressed AI Models
With the funding, Multiverse’s platform will grow, and compressed AI models will be available on more cloud platforms and in more industries. Additionally, the company intends to expand its model library and improve its approach, which is influenced by quantum computing. This will assist developers in industries like finance, healthcare, and logistics who use open-source AI tools.
Thus, Multiverse’s innovation may establish new standards for the application of AI, given the growing demand for effective language models. Furthermore, the technology supports international initiatives to lower the operational and environmental costs of AI. Therefore, compression might be necessary to maintain speed, affordability, and energy efficiency as larger models continue to dominate the market.
Compressed AI Models Signal a New Era
Multiverse Computing is transforming the way AI models are developed and deployed. Their ability to reduce size without compromising output is opening the door to a new generation of affordable and accessible AI. Additionally, their work creates a distinct edge in a growing market by fusing the accessibility of open-source AI with the potential of quantum computing. The startup’s recent funding solidifies its position as a major player in the European tech industry. As more businesses use AI models, effective solutions like Multiverse’s will probably become the norm for successful, scalable AI.
About The Author
