Nvidia unveils 120B-parameter Nemotron Super VL

Nvidia on Thursday launched open-weight artificial intelligence models for enterprise agents and local computing, including the 120-billion-parameter Nemotron 3.5 Super VL and two smaller Bolt models.

Super VL and Nemotron 3.5 Bolt 74B-A9B 24GB are available immediately, while Bolt E4B is expected later this month, Nvidia said.

Super VL is designed for enterprise agents. Nvidia said it delivers instruction-following and tool-use accuracy similar to Nemotron 3 Ultra despite having fewer parameters.

Nvidia claims Super VL completes specialized agent tasks faster than competing Anthropic and OpenAI models, including Opus 5, GPT 6.1 Sol, Luna and Astra, after post-training for those tasks.

The tasks include document intelligence, financial accounting, scientific experimentation, medical insurance claims and security attack resilience.

Nvidia said companies using Super VL include Skydance, EY, Ramp, Crowdstrike, Perplexity and Salesforce.

“Any organization that wants to run agents on Nvidia infrastructure now has an open, optimized model for their tier without having to choose between capability and deployability,” Kari Briski, Nvidia’s vice president of Open Models and Software, wrote in a blog post announcing the models.

Super VL joins Nvidia’s Nemotron family of open-weight models, which covers multimodal AI, reasoning, coding, mathematics and enterprise automation.

Proprietary frontier AI developers, which are both Nvidia customers and recipients of its investments, are looking beyond its hardware to alternatives including AWS Trainium, Cerebras systems and chips they develop themselves.

Nvidia is targeting enterprises that want to run their own open models in their own data centers.

“Unlike closed frontier models, organizations can post-train Super VL on their own data, deploy it on their own infrastructure and own what they build,” Briski wrote.

The two Bolt models are intended to run locally on AI PCs using Nvidia’s RTX Spark chip and on DGX Spark workstations. The PCs, aimed at consumers, programmers and enterprise knowledge workers, are entering the market through partners including Microsoft and Dell.

Nvidia also hopes local models and agents will create demand for PCs powered by its chips. Briski wrote that PC users “can delegate everyday projects … without sending those workloads to the cloud.”

Featured image credit