Multiverse Computing is pioneering a new generation of model compression tools that can shrink large language models (LLMs) by up to 95%, while maintaining performance and slashing costs. With its latest €189 million (~$217M) funding round, the company has emerged as Spain’s largest AI startup.

By applying quantum-inspired machine learning, Multiverse enables efficient deployment of LLMs like Llama, DeepSeek, and Mistral—without needing quantum hardware. These compressed models reduce cloud computing costs by up to 80%.

Its tools are now available through the AWS AI Marketplace, helping businesses worldwide scale AI faster, cheaper, and more sustainably.

Key Features:

• Up to 95% Model Compression: Reduces the size of LLMs with minimal performance loss.

• 80% Cost Reduction: Dramatically cuts GPU and cloud costs for enterprises.

• No Quantum Hardware Required: Uses quantum-inspired methods on classical systems.

• Global Reach: Deployed via AWS Marketplace for easy enterprise adoption.

Source:
Reuters — Full Article