Nvidia has launched new open-weight models, Nemotron 3.5 Super VL and two Bolt models, designed for always-on agents and edge computing. The Nemotron 3.5 Super VL is a 120-billion-parameter model aimed at enterprise agents, which Nvidia states can deliver similar accuracy to its Nemotron 3 Ultra model in instruction-following and tool-use tasks.
The company also introduced two lightweight Bolt models intended to run locally on new AI PCs, which feature the Nvidia RTX Spark chip, and workstations running the DGX Spark. These PCs are being released into the market through partners such as Microsoft and Dell, targeting consumers, programmers, and enterprise knowledge workers.
Nvidia claims the Super VL completes tasks faster than some competing models when post-trained on specific agentic tasks, including document intelligence and financial accounting. Companies reportedly using Super VL include Skydance, EY, Ramp, Crowdstrike, Perplexity, and Salesforce. The Super VL and the 74-billion-parameter Bolt 74B-A9B 24GB are available now, with the smaller Bolt E4B model expected later this month.