Overview
multiverse-computing platform applying tensor-networks to LLM model-compression. Includes CompactifAI Router for real-time local vs cloud routing. Marketed for sovereign/edge/on-prem efficient inference.
Key Claims
Warning
Up to 80–95% size reduction with “immaterial” accuracy loss is a vendor performance claim. Academic CompactifAI work showed large compression on LLaMA-2 7B with quantization + healing, with small accuracy drops — not a substitute for independent product audits.