131 likesNew · 4d old
8B Chat Model Fits in 4 GB, Targets Edge Assistants
What is NVIDIA-Nemotron-3.5-Bolt-E4B-GGUF?
NVIDIA’s Nemotron-3.5 Bolt-E4B is a 4-billion-effective-parameter chat model that runs in just 4 GB of RAM thanks to Q4_K_M quantization. It’s optimized for fast, multi-turn conversations on local voice assistants and embedded AI interfaces.
Who it is for
If you build on-device bots or interactive personas, you can now run a capable LLM locally without relying on cloud services.