Multiverse Computing Pushes Its Compressed AI Models Into the Mainstream

Why It Matters

On‑device AI reduces reliance on costly, volatile cloud compute while enhancing data privacy, a growing priority as AI adoption scales across regulated and mission‑critical sectors.

Key Takeaways

• CompactifAI runs Gilda locally on compatible devices
• API portal offers self‑serve access to compressed models
• HyperNova 60B cuts latency and cost versus original LLM
• Over 100 enterprise customers include Bank of Canada, Bosch
• Multiverse raised $215M Series B, eyeing €500M round

Pulse Analysis

Rising AI compute costs and tightening venture capital terms have pushed firms to reconsider the traditional cloud‑first model. Multiverse Computing’s quantum‑inspired compression technology enables models like Gilda to fit on smartphones, delivering true offline interaction and eliminating third‑party data exposure. This edge‑first approach aligns with a broader industry shift toward localized inference, where latency, bandwidth constraints, and regulatory privacy mandates drive demand for on‑device solutions.

For enterprises, the new CompactifAI API portal removes the friction of marketplace intermediaries, offering transparent usage metrics and direct deployment of compressed models. Compared with competitors such as Mistral Small 4 or Apple Intelligence’s hybrid strategy, Multiverse’s HyperNova 60B delivers comparable performance at a fraction of the compute budget, making it attractive for high‑volume, agentic workloads like autonomous coding assistants. The self‑serve model also empowers developers to fine‑tune cost and latency trade‑offs without sacrificing observability, a critical factor for production‑grade AI pipelines.

Beyond cost savings, edge‑ready models unlock use cases where connectivity is unreliable—drones, satellites, field equipment, and secure government environments. Multiverse’s existing roster, including the Bank of Canada, Bosch, and Iberdrola, demonstrates early traction in sectors that value data sovereignty and resilience. Backed by a $215 million Series B and a potential €500 million raise, the startup is positioned to capitalize on the growing market for compact, privacy‑preserving AI, signaling a maturing ecosystem where on‑device intelligence becomes a mainstream enterprise capability.