HyperNova 60B 2605

ResearchOpen weightsText60B parameters

HyperNova family model by CompactifAI.

About the HyperNova family

Multiverse Computing's compressed-model line — tensor-network-compressed LLMs (e.g., HyperNova 60B) delivering near-original quality in a fraction of the memory.

All HyperNova models

How this one is configured

This configuration ~60B parameters.

Who makes it

CompactifAISpainQuantum-inspired compression making big models radically smaller.

← Back to all models