NVIDIA releases Nemotron 3.5 Bolt, a 24GB vision-language model for local agent tasks
Original titlenvidia/NVIDIA-Nemotron-3.5-Bolt-74B-A9B-24GB
AISummary
NVIDIA releases Nemotron 3.5 Bolt 74B-A9B 24GB, a vision-language model that uses 2-bit quantization to fit in 24GB of memory. The model card says it targets local agentic tasks and reports benchmark results against Qwen3.6-35B-A3B, Gemma4 26B-A4B, and Nemotron 3.5 Lightning 30B-A3B. It supports up to 256K tokens of context and is released under the OpenMDW-1.1 license.
AIWhy it matters
The model card details how a 74B vision-language model was compressed to 24GB, a useful trade-off reference for teams planning local deployment.
Source: NVIDIA · new models on Hugging Face · huggingface.coPublished · added here