Spanish AI Firm Multiverse Computing Targets €500M Funding Round at €1.5 Billion Valuation

Multiverse Computing Eyes Major Funding Round

Multiverse Computing is in advanced discussions to raise approximately €500 million funding in a round expected to value the Spanish artificial intelligence software company at more than €1.5 billion, according to sources familiar with the matter 1

1

. The financing, anticipated to close during the first half of 2026, will bring in new backers alongside existing investors including HP Tech Ventures, Forgepoint Capital, and CDP Venture Capital 1

1

. This development comes roughly eight months after the San Sebastian-based company raised $215 million from a consortium that included Toshiba Corp. and other major players 2

2

.

The funding round would elevate Multiverse Computing into Europe's expanding cohort of AI unicorn startups valued above $1 billion. The company's annual recurring revenue reached €100 million in January, demonstrating significant traction for its approach to making AI deployment more cost-effective for enterprises 1

1

.

AI Model Compression Technology Drives Growth

Founded in 2019, Multiverse Computing has built its business around addressing one of AI's most pressing challenges: the enormous computational and energy requirements of modern AI systems. The company's flagship product, CompactifAI, tackles this by reducing the hardware footprint needed to run AI models through an innovative approach that converts weight matrices into tensor networks 2

2

. According to the company, this AI model compression technology can halve training times and accelerate inference by 25% while simultaneously reducing storage requirements .

The process of optimizing large language models introduces some errors, which CompactifAI mitigates through a technique the company calls "healing" - retraining the neural network after compression using only a handful of graphics cards. In a May 2024 research paper, Multiverse demonstrated that it successfully compressed the Llama 2 7B language model by 93% in memory footprint, with only a 3% drop in output accuracy when combined with quantization methods 2

2

.

Market Positioning and Customer Adoption

Multiverse Computing reduces hardware footprint requirements at a time when enterprises face mounting pressure to deploy AI efficiently. The company serves more than 100 organizations, including major enterprises like Allianz SE and Moody's Corp. 2

2

. Customers can either install CompactifAI on their own infrastructure or access pre-compressed open-source AI models through an application programming interface.

Beyond CompactifAI, the company offers the Singularity platform, which provides access to AI models designed for industry-specific use cases across multiple verticals. These modules range from detecting factory equipment malfunctions to helping financial professionals make investment decisions and identifying malicious network traffic in cybersecurity applications 2

2

.

Quantum Computing Ambitions

Multiverse Computing is also positioning itself to capitalize on emerging interest in quantum computing. The startup's team includes computational scientists and quantum physicists who have partnered with major quantum hardware manufacturers for software development 1

1

. While useful quantum computers and applications are thought to be at least five to 10 years away, this strategic positioning could provide long-term growth opportunities as the technology matures.

The company is banking on growing demand driven by the construction of new AI "gigafactories" in Europe and the implementation of AI technology on devices ranging from satellites and drones to smartphones 1

1

. For investors unable to back US giants like OpenAI, European teams building credible models and software businesses represent an attractive alternative for capturing returns in the AI boom.