NVIDIA Nemotron 3.5 Super VL: Open Multimodal Model Built for Agents
Melvin Vivas · X post · 2026-10-15 · 1:20 · 446 views · Open on X
Topics: Industry Trends & Job Market, LLM Fundamentals, AI Agents, Tool Use & MCP · Level: intermediate
Summary
This post shares NVIDIA's announcement of Nemotron 3.5 Super VL. It is an open model that combines reasoning and tool use with understanding of text, images, documents and video, and it is meant for long-running AI agents. It is a mixture-of-experts model with 120B total parameters, of which 12B are active.
Key points
- Nemotron 3.5 Super VL is an open model from NVIDIA.
- It combines reasoning and tool use in one model.
- It handles text, image, document and video inputs.
- It is aimed at long-running AI agents.
- It has 120B total parameters but only 12B are active per token, a mixture-of-experts design that keeps inference cost down.
Resources mentioned
- NVIDIA Nemotron 3.5 Super VL · tool · huggingface.co · free
NVIDIA's open 120B-A12B multimodal model with 1M context, built for agents working across text, images, documents, GUIs and video.
Also in: NVIDIA Nemotron 3.5 Super VL now on OpenRouter (Melvin Vivas on X · notes)
More in Industry Trends & Job Market
- FLUX 3 Image Klein: Open-Weights Image Model Now Supported in ComfyUI
- Krea 3 Image Editing Model Now Available on fal
- Krea 3: Fast, Precise AI Image Editing Model Enters Public Beta
- Clueso's $10M Series A: AI Automation for Company Video Creation
- Gemini 4 Argon reportedly close to Claude Opus 5.5
- Fin Apex 2.0: Customer-Service Models Built on NVIDIA Nemotron 3 Super