Multiverse Computing debuts the CompactifAI App, a mobile solution that runs advanced AI models locally on devices offline or switches to cloud-based models via API, with a focus on edge AI.

The app enables fully offline edge operation, smart routing between on-device and API models to balance speed and capability, and privacy-by-design with local data processing.

This approach supports privacy-centered security and reduces reliance on cloud infrastructure for sophisticated reasoning.

Details on pricing, monetization, and initial customer adoption remain limited, making near-term financial impact uncertain and dependent on execution.

Multiverse Computing is a Donostia, Spain–based company with offices in the US, Canada, and Europe, serving more than 100 global customers including Iberdrola, Bosch, and the Bank of Canada.

Headquartered in Donostia, the company maintains a multinational presence and a sizable enterprise customer base across industries.

Multiverse showcased CompactifAI at MWC26 in the Spanish Pavilion, Hall 4, signaling efforts to broaden its commercial reach in edge AI and LLM tooling.

The launch follows the earlier release of HyperNova, an open-source compressed model, extending efficiency gains from research to real-world deployments.

HyperNova represents a stepping stone in Multiverse’s trajectory toward practical, efficient AI for deployment.

The LinkedIn post highlights privacy and security benefits, targeting users in regulated or data-sensitive sectors and scenarios with low connectivity.

CompactifAI uses quantum-inspired mathematics to compress AI models by up to 95% with only a 2-3% precision loss, outperforming typical compression trades.

The technology leverages Multiverse's CompactifAI approach to deliver high compression with minimal accuracy sacrifice, distinguishing it from standard methods.