Model5d ago
Gemma 4 Debuts On-Device Multimodal AI
Gemma 4, a 12‑billion‑parameter multimodal model, launches on-device. Built by Hugging Face, it runs on mobile GPUs, handling text, image, and audio in a single inference. Developers can embed full multimodal intelligence locally, cutting latency and avoiding cloud dependencies. The release shows on‑device AI can match cloud‑scale performance for enterprise.