---
title: "Nvidia Releases Open-Weight Model Nemotron 3.5 Super VL"
url: https://newscentral.site/nvidia-releases-open-weight-model-nemotron-3-5-super-vl/
language: en
publisher: "News Central Site"
section: "Technology"
published: 2026-10-15T15:00:00.000Z
updated: 2026-10-15T23:57:24.281Z
id: f069c8eb-5ce8-4aee-83f2-092ced290ee0
source: "ZDNet https://www.zdnet.com/innovation/nvidia-nemotron-3-5-super-vl-model-us-open-weight-ai/"
attribution: "Link to https://newscentral.site/nvidia-releases-open-weight-model-nemotron-3-5-super-vl/ and name News Central Site when you quote or summarize this story."
---

# Nvidia Releases Open-Weight Model Nemotron 3.5 Super VL

Nvidia's latest open-weight model, the Nemotron 3.5 Super VL, boasts an impressive 120 billion parameters and delivers high-speed task completion and accuracy for specialized agents.

A significant shift is underway in the field of open-weight AI, where Chinese labs have traditionally held a strong advantage. However, recent years have seen a surge in open models developed by Western companies, challenging this dominance.

Nvidia has released its latest offering, the Nemotron 3.5 Super VL, an open-weight model aimed at addressing the growing demand for agentic work. This new model boasts an impressive 120 billion parameters and is designed to deliver high-speed task completion and accuracy for specialized agents.

According to Nvidia's claims, the Nemotron 3.5 Super VL outperforms its competitors in terms of speed, achieving a significant milestone as the fastest open agentic model in its class. In benchmark testing conducted by Artificial Analysis, the model demonstrated exceptional performance, leaving other similarly sized models like Ling 3.0 Flash and V4.1 Flash far behind.

Nvidia has emphasized the capabilities of the Nemotron 3.5 Super VL to achieve top-tier accuracy for specialized agents through post-training. The company highlighted a notable example where the base model saw a substantial gain in medical insurance claims, reaching a 44% improvement when fine-tuned.

The release of this new model provides enterprises with an opportunity to tailor their AI solutions and reap specific benefits, as seen in the case study mentioned by Nvidia. While proprietary models may still hold a slight edge, Super VL's adaptability gives it a distinct advantage in this competitive landscape.

The Nemotron 3.5 Super VL from Nvidia offers businesses greater flexibility and control over their AI models by allowing them to post-train on their own data and run on their own infrastructure.

This adaptability is a significant advantage in the competitive landscape of AI, where proprietary models often hold a slight edge due to their tailored performance and support.

Nvidia's new edge models, Nemotron 3.5 Bolt 74B-A9B 24GB and Nemotron 3.5 Bolt E4B 4GB, are designed for local operation and off-grid use, catering to different needs and applications.

The company claims that these edge models deliver exceptional intelligence per gigabyte of memory, making them suitable for various tasks such as voice recognition, device agents, and interactive characters.

While the efficiency gains demonstrated by Artificial Analysis' findings suggest a promising future for Nvidia's new model, it remains to be seen whether it will set a new standard for open-weight or proprietary models in practice.

The release is part of a broader shift in the AI world, where open models from companies like DeepSeek and Moonshot.ai are gaining traction due to their promise of customization, cost-effectiveness, and flexibility.

The shift towards open-weight AI models is gaining momentum in the industry, with US-made models now being considered a viable option for various applications. This change in perception follows the OpenAI Hugging Face breach, which challenged the notion that only open models pose security risks.

This development may help close the geography gap between US and Chinese open-weight models, as American companies like Nvidia are now competing directly with their Chinese counterparts. As a result, the use of foreign technology is being reevaluated in light of cybersecurity guidelines, potentially paving the way for increased adoption of open models globally.

---
Source: [ZDNet](https://www.zdnet.com/innovation/nvidia-nemotron-3-5-super-vl-model-us-open-weight-ai/)  
Published by News Central Site: https://newscentral.site/nvidia-releases-open-weight-model-nemotron-3-5-super-vl/