Spanish startup Multiverse Computing, specializing in quantum and tensor technologies for artificial intelligence, has announced the closing of a Series C funding round of up to $570 million. The company's valuation has reached $1.7 billion, confirming the growing market interest in solutions that optimize the resource intensity of large language models.
The startup's key product, the CompactifAI platform, uses the mathematical framework of tensor networks to compress AI models. According to stated data, the technology can reduce model size by up to 95% without noticeable loss of accuracy or performance. This is especially relevant in a context where infrastructure costs for deploying and maintaining large neural networks are becoming one of the main barriers to widespread AI adoption.
Where the funds will go
The raised capital will be directed toward scaling the CompactifAI platform and creating so-called "sovereign AI infrastructure"—solutions that allow companies and governments to deploy their own computing capabilities without dependence on external cloud providers. Additionally, Multiverse Computing plans active expansion into the markets of the United States, Asia, and the Middle East, where demand for efficient and energy-saving AI solutions is growing exponentially.
Cryptalist expert opinion: The trend toward "lightweight AI" is not just a fad but a necessity. When the cost of training a single model can exceed tens of millions of dollars and data center energy consumption becomes a critical factor, solutions like CompactifAI change the game. If the technology truly preserves quality at such a level of compression, we are witnessing the birth of a new standard for the enterprise sector and government agencies. Multiverse Computing is betting that the future of AI lies not in endlessly increasing parameters, but in smart optimization.