ITIT Solution JournalEnterprise IT, reported straight

NVIDIA Releases Nemotron 3.5 Super VL and Two Edge Models

Three new open Nemotron models target enterprise servers, local workstations, and edge devices.

IT Solution Journal News Desk

Editorial illustration: Three matte black egg-shaped nesting forms of decreasing size sit open on a dark surface, the largest revealing a blue-lit server rack interior, the middle a white-lit miniature workstation with monitors, and the smallest a warm amber-lit tiny chest, each glowing with its respective color onto the floor.

AI-generated editorial illustration.

NVIDIA on October 15 released Nemotron 3.5 Super VL, a 120-billion-parameter open model the company says delivers frontier accuracy on multimodal agentic tasks at roughly one-fifth the parameter count of its largest reasoning model. Two edge-focused Bolt models accompany it: Bolt 74B-A9B 24GB, available now for local agentic workloads on RTX and DGX Spark hardware, and Bolt E4B 4GB, arriving October 23 for conversational and embedded agents on GeForce RTX and Jetson devices. NVIDIA says Super VL achieves up to 5x faster throughput than comparable models in its class. All three ship with open weights. Adopters cited include CrowdStrike for security, Salesforce for government workflows, and Atlassian for agentic search.

Source: NVIDIA News. Original announcement: October 15, 2026.

Keep up with IT Solution Journal

Practical guidance and new coverage. You can withdraw your permission at any time.

Read our privacy and data-use policy.