Nvidia has launched new open-weight Nemotron 3.5 models designed for always-on AI agents, enterprise workloads, and local AI PCs.
The lineup includes the 120-billion-parameter Nemotron 3.5 Super VL for enterprise agents, along with two smaller Bolt models intended to run locally on Nvidia-powered PCs and workstations, including new RTX Spark-powered laptops.
Nemotron 3.5 Super VL
Nemotron 3.5 Super VL uses fewer parameters than Nvidia’s Nemotron 3 Ultra but is designed to deliver similar accuracy in instruction-following and tool-use tasks.
Nvidia claims Super VL can complete certain agentic tasks faster than competing models including Claude Opus 5, GPT-6.1 Sol, Luna and Astra after being post-trained for specific workloads.
Those workloads include document intelligence, financial accounting, scientific experiments, medical insurance claims, and resilience against security attacks.
Nvidia says companies including Skydance, EY, Ramp, CrowdStrike, Perplexity and Salesforce are already using Super VL.
The company is positioning the model as an option for businesses that want to post-train AI on their own data and run it on their own infrastructure instead of depending on closed frontier models.
Smaller Bolt Models Run Locally
Nvidia also introduced two smaller models aimed at local AI agents.
Nemotron 3.5 Bolt 74B-A9B 24GB is a 74-billion-parameter model designed for local agentic workloads. It targets tasks including multistep reasoning, tool calling and web browsing while balancing speed and reasoning ability.
The smaller Bolt E4B is aimed primarily at conversational AI agents.
These models are intended for new AI PCs powered by Nvidia’s RTX Spark chip and workstations using DGX Spark.
Nvidia says running agents locally could let users handle everyday tasks without sending sensitive workloads involving areas such as health, finances or family information to the cloud.
Nvidia Pushes Further Into Open Models
The new releases expand Nvidia’s existing Nemotron family, which already covers multimodal AI, reasoning, coding, mathematics and enterprise automation.
The move comes as major AI companies increasingly explore hardware beyond Nvidia, including AWS Trainium and Cerebras, while also developing their own chips.
At the same time, Nvidia is targeting growing demand from companies that want to operate open models inside their own data centers.
Nemotron 3.5 Super VL and Bolt 74B-A9B are available now. Nvidia says Bolt E4B will arrive later this month.
The post Nvidia Launches New Open Model for Powerful AI Agents That Run Locally appeared first on ProPakistani.