Nvidia Launches New Open-Weight AI Models for Enterprise and PC Use
Nvidia has introduced new open-weight models tailored for enterprise agents and edge computing. The Nemotron 3.5 Super VL, a large model with 120 billion parameters, is designed to deliver high accuracy in tasks like document intelligence and financial accounting while being more efficient than previous versions. Alongside this, Nvidia also launched two lightweight Bolt models intended for local use on AI PCs powered by the RTX Spark chip and workstations using DGX Spark hardware. The company claims that Super VL outperforms competing models from Anthropic and OpenAI in specific agentic tasks when post-trained. Notable companies like Skydance, EY, and Salesforce are reportedly already utilizing it. Nvidia is positioning these open-weight models as a way for organizations to maintain control over their AI infrastructure without relying on proprietary frontier labs. In addition to enterprise applications, the Bolt models aim to support consumer-level AI assistants by enabling local processing of tasks like web browsing and tool calling, addressing privacy concerns associated with cloud-based solutions. The larger Bolt model is available now while a smaller version will be released later this month.
Tickers: $CRM $CRWD $MSFT $NVDA
Topics: AI edge computing enterprise agents Nemotron models open-weight AI
Read the full article at TechCrunch
Part of the story Meta Launches Open-Weight AI Model Optimized for Macs (3 articles)