LLM
NVIDIA Nemotron 3.5 Bolt 74B-A9B 24GB
NVIDIA Nemotron 3.5 Bolt 74B-A9B 24GBA vision-language model with 74B total parameters and 9B active, quantized for a 24GB memory target. Built for agent workflows using text, images, video, and tools.
Model details
View repositoryNVIDIA Nemotron 3.5 Bolt 74B-A9B 24GB is an open-weight vision-language model designed for local agent workloads. It accepts text, images, and video, and generates text. The model has 74B total parameters, 9B active parameters, and a hybrid Mamba-2–Transformer MoE architecture with a C-RADIO v4-H vision encoder. IQ2 quantization is designed to fit the model within a 24GB memory budget.
Use it for agent workflows that combine visual understanding with tool use, web browsing, and terminal tasks.
