The inherent scale of large language models (LLMs) presents a significant challenge for widespread adoption. Addressing this, Spanish startup Multiverse Computing is pioneering the development of compressed models designed to bridge the gap between the advanced capabilities of frontier models and the practical deployment affordability for businesses.
At the core of Multiverse's innovation is CompactifAI, a proprietary compression technology drawing inspiration from quantum computing principles. This technology, developed by the Basque-based company, has been successfully applied to models originally released by OpenAI. Developers can now freely access an updated version of Multiverse’s HyperNova 60B model on Hugging Face. Furthermore, the company has outlined plans to open-source additional compressed models in 2026, aiming to cater to an even broader spectrum of use cases.
Multiverse asserts that its models, despite their reduced size, maintain comparable potency and accuracy to their larger counterparts. HyperNova 60B, for instance, occupies just 32GB, approximately half the footprint of its source model, OpenAI’s gpt-oss-120B, while also delivering reduced memory consumption and lower latency. The latest iteration, HyperNova 60B 2602, introduces enhanced support for tool calling and agentic coding, areas typically associated with high inference costs.
Notably, Multiverse claims that HyperNova 60B has outperformed Mistral Large 3, a prominent model from the French AI decacorn, Mistral AI. Beyond this technological rivalry, both European AI entities share several significant characteristics, underscoring a broader trend in the continent's AI landscape.
Mirroring Mistral's global expansion, Multiverse has established an international presence with offices spanning the United States, Canada, and various European locations. Both companies also boast a robust portfolio of enterprise clients. Multiverse’s notable customers include Iberdrola, Bosch, and the Bank of Canada, highlighting its reach across diverse sectors.
While Multiverse has yet to officially achieve unicorn status, there are current rumors circulating about a prospective €500 million funding round, potentially valuing the company at over €1.5 billion. In a statement provided to TechCrunch, Multiverse confirmed ongoing active discussions with potential investors regarding a new funding round but cautioned that commenting on specific valuation or funding size at this juncture would be premature. The company also chose not to address reports indicating its annual recurring revenue (ARR) reached €100 million in January.
Should the reported €100 million ARR be confirmed, it would represent a fraction of OpenAI’s $20 billion ARR, yet it would position Multiverse closer to Mistral’s ARR, which surged past $400 million, partly fueled by a growing demand for alternatives to U.S.-based technology. This aligns with Multiverse’s strategic positioning, as articulated in its recent press release, to “deliver sovereign solutions across the AI stack.”
These geopolitical considerations have recently facilitated a significant collaboration between Multiverse and the regional government of Aragón in northeastern Spain. Furthermore, the Spanish Agency for Technological Transformation (SETT) was a participant in the AI startup’s $215 million Series B funding round last year. Since its inception, Multiverse has consistently received support from the Basque region, which could soon celebrate the emergence of its first unicorn.
The Editorial Staff at AIChief is a team of professional content writers with extensive experience in AI and marketing. Founded in 2025, AIChief has quickly grown into the largest free AI resource hub in the industry.