Multiverse Computing Lands $215M to Drive Next-Gen AI Efficiency

Spanish startup Multiverse Computing has secured a significant €189 million (approximately $215 million) in Series B funding to expand its innovative quantum-inspired technology, CompactifAI, which promises to dramatically reduce the size and cost of running large language models (LLMs).

Shrinking AI Models with Quantum-Inspired Compression

At the heart of Multiverse Computing’s breakthrough is CompactifAI, a compression method influenced by quantum computing principles. According to the company, this technology can shrink the size of LLMs by up to 95% — all without any drop in model performance.

Instead of creating entirely new AI models, Multiverse focuses on offering compressed versions of existing open-source LLMs. These include:

• Llama 4 Scout

• Llama 3.3 70B

• Llama 3.1 8B

• Mistral Small 3.1

The company also plans to release a compressed version of DeepSeek R1 soon and is currently working on additional open-source and reasoning models. However, proprietary models, such as those developed by OpenAI, are not supported at this time.

Faster, Cheaper AI with “Slim” Models

Multiverse refers to its compressed versions as “slim” models, which are available on Amazon Web Services (AWS) or through on-premises licensing. These slim models are significantly faster — 4 to 12 times — compared to their uncompressed counterparts. That speed translates into meaningful savings: inference costs are reduced by 50% to 80%.

For example, Llama 4 Scout Slim costs just $0.10 per million tokens on AWS, compared to $0.14 for the original version.

Beyond cloud usage, Multiverse’s compact models are so efficient that they can run on PCs, smartphones, cars, drones, and even devices as small as a Raspberry Pi — offering low-power AI capabilities at the edge.

Founders with Deep Technical and Financial Expertise

Multiverse was co-founded by CTO Román Orús, a professor at the Donostia International Physics Center in San Sebastián, Spain. Orús is recognized for his pioneering research in tensor networks — mathematical structures that simulate quantum computing capabilities using classical machines. These tools are now being leveraged to compress AI models with high efficiency.

CEO Enrique Lizaso Olmos, the company’s other co-founder, holds several mathematics degrees and has a background in academia and finance. He is a former deputy CEO at Unnim Banc and brings both analytical and business acumen to the startup’s leadership.

Backed by Top-Tier Investors

Multiverse’s Series B funding was led by Bullhound Capital, known for backing high-profile tech companies like Spotify, Revolut, Delivery Hero, Avito, and Discord. The round also included investments from:

• HP Tech Ventures

• SETT

• Forgepoint Capital International

• CDP Venture Capital

• Santander Climate VC

• Toshiba

• Capital Riesgo de Euskadi – Grupo SPR

Global Reach and Future Growth

With 160 patents and a customer base that includes major players like Iberdrola, Bosch, and the Bank of Canada, Multiverse is making its mark on the global stage. The new funding brings its total capital raised to approximately $250 million, reinforcing its position as a frontrunner in AI efficiency technology.

For companies looking to scale AI affordably and sustainably, Multiverse’s “slim” approach might just reshape the future of intelligent computing.

Author: Din Kumar