NVIDIA Nemotron 3.5 Bolt E4B is a small language model for local conversational agents, running in ~4GB VRAM.