Nvidia introduces new open models to enable enterprises and PCs to run agents.

Nvidia on Thursday launched new open-weight models designed for always-on agents and edge computing, with Nemotron 3.5 Super VL at the center. The release puts the company's AI stack into two connected arenas: enterprise agents that need accuracy and local machines that must keep workloads away from the cloud.

Background: A Wider Open-Model Race

Nvidia's move builds on an existing Nemotron family that already covers multimodal artificial intelligence, reasoning, coding, math and enterprise automation. These open-weight models compete with proprietary frontier labs, while those labs are also Nvidia customers and investments.

The competitive frame has widened. Fast-growing proprietary AI labs are now looking to many other chips and AI system makers beyond Nvidia to power their needs, including AWS's Trainium and Cerebras, as well as building their own home-grown chips. Nvidia's response in this announcement is to lean hard into enterprise companies wanting 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," Kari Briski, Nvidia's vice president of Open Models and Software, wrote in the blog post announcing these models.

Development: Enterprise Agents and Local AI PCs

The main enterprise model is Nemotron 3.5 Super VL, a large, 120-billion-parameter model geared for agents used by the enterprise. Nvidia says it uses fewer parameters than the Nemotron 3 Ultra model while delivering similar accuracy in instruction-following and tool-use tasks.

At the lighter end, Nvidia launched two Bolt models for local use. Nemotron 3.5 Bolt 74B-A9B 24GB is a 74-billion-parameter model designed for local agentic tasks, intended to balance speed with reasoning to handle multistep reasoning, tool calling and web browsing. The smallest Bolt E4B is aimed at conversational agents.

These smaller models are intended to run locally on new AI PCs running the Nvidia RTX Spark chip and workstations running the DGX Spark. Such PCs, aimed at consumers, programmers and enterprise knowledge workers, are being released right now into the market through partners like Microsoft and Dell.

"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," Briski wrote.

Latest Progress: Claims, Adoption and Availability

Nvidia claims that Super VL completes tasks faster than competing models from Anthropic and OpenAI (Opus 5, GPT 6.1 Sol, Luna, Astra) when post-trained on specific agentic tasks. The tasks listed include document intelligence, financial accounting, scientific experimentation, medical insurance claims, and security attack resilience.

The company also backed its claims with a list of heavy-hitter companies it says are using Super VL: Skydance, EY, Ramp, Crowdstrike, Perplexity and Salesforce. Such endorsements give the launch a concrete adoption signal, but the timing claim remains tied to post-training on the named agentic tasks.

The consumer angle is tied to a nascent trend of AI assistants. A new AI assistant seems to be introduced every day by startups and big tech alike, while privacy weary consumers may not want to invite such agents deeply into their lives, giving them access to information about their families, health, finances and so on. Nvidia hopes to create a new market for itself selling PCs powered by its chips by creating machines that run the models and agents locally.

PC users "can delegate everyday projects … without sending those workloads to the cloud," Briski wrote.

Super VL and Bolt 74B-A9B are available now, with Bolt E4B coming later this month, Nvidia says. The immediate watch point is Bolt E4B's arrival, while the broader one is whether the AI PCs released through Microsoft and Dell can bring these local models to consumers, programmers and enterprise knowledge workers.

Meta description: Nvidia releases Nemotron 3.5 Super VL, a 120 billion parameter open-weight model, and Bolt models for local AI PCs, agents and enterprise workloads.

Tags: Nvidia, Nemotron 3.5 Super VL, AI PCs, open-weight models, enterprise agents

Featured image: abstract circuit patterns around a glowing AI chip, technology style, no people, no logos, no text