Key Info

NVIDIA Nemotron 3.5 Super VL, an open 120B-A12B multimodal model from NVIDIA AI, is now live on OpenRouter with a 1M token context window.

Highlights

  • Open 120B-A12B multimodal model designed for long-running agents that see, reason, and act
  • Handles text, images, documents, GUIs, and video
  • Hybrid Mamba-Transformer MoE architecture with multi-token prediction for faster output
  • Targeted at document intelligence, web research, customer service, and video understanding
  • Available to try now on OpenRouter