Nvidia Unleashes Nemotron 3.5: Faster Open AI Agents Emerge

Nvidia has introduced its Nemotron 3.5 Super VL, a new 120-billion-parameter open-weight AI model designed for enhanced efficiency and speed in specialized agentic work. This model is positioned to accelerate task completion, with Nvidia claiming it is five times faster than its nearest competitors. The introduction of Nemotron 3.5 Super VL marks a significant development in the competitive landscape of open AI models.

This release is part of a growing trend of Western AI labs developing advanced open-weight models. Previously, Chinese AI companies largely led this space. However, recent contributions from labs like France's Mistral and Nvidia-backed Reflection AI indicate a closing geographical gap. The industry's increased embrace of open models has been partly fueled by security concerns previously associated only with open models.

Nvidia also unveiled two edge-focused models, Nemotron 3.5 Bolt, designed for local agentic use and fast conversational applications. These models can be customized by enterprises using their own data and infrastructure, offering enhanced personalization and security compared to proprietary alternatives. This approach allows for significant performance gains in specific domains, such as a cited 44% improvement in medical insurance claims processing.