← All model drops
Live drop
Nemotron 3.5 Super VL
NVIDIADropped Oct 15, 2026
Open multimodal MoE for always-on agents with text/image/video in one set of weights.
About this release
Live drop
NVIDIADropped Oct 15, 2026
Open multimodal MoE for always-on agents with text/image/video in one set of weights.
About this release
Nemotron 3.5 Super VL is an open 120B-parameter (12B active) Mixture-of-Experts multimodal model from NVIDIA, released October 15, 2026. It handles instruction following, tool use, multistep workflows, and visual understanding across text, images, documents, and video in a single model. Weights, datasets, and training recipes are available on Hugging Face under an open license for customization and deployment across data center, edge, and consumer hardware.
Read the official NVIDIA launch pageFirst builds incoming — this lane fills in automatically as people start building with Nemotron 3.5 Super VL. Know one? Submit the URL.