Multiverse Computing Releases LittleLamb: Ultra-Compact 0.3B AI Models for Edge Devices
Multiverse launched three LittleLamb models on Hugging Face: general, tool-calling, and mobile variants, compressed 50% from Qwen3-0.6B for bilingual edge AI. They outperform peers on benchmarks with dual thinking/non-thinking modes for low-latency tasks. Targets mobile, offline, and agentic apps.