NVIDIA released Nemotron 3.5 Bolt E4B 4GB, a small language model post-trained from Gemma 4 E4B. The company targets it at local voice assistants, game characters and robots, with about 4GB of VRAM on RTX and Jetson devices. The source does not give performance results, licensing details or specific availability information.
Why it matters
Developers working on local voice assistants, game characters or robots have a model NVIDIA says targets modest device memory, but the source does not establish its performance or licensing terms.
