NVIDIA said in a blog post that it released Nemotron 3.5 Super VL, an open 120-billion-parameter model that it says delivers the fastest task completion and highest accuracy for specialized always-on agents.[1] Super VL and Bolt 74B-A9B are available now, with Bolt E4B coming in October, according to NVIDIA.[1] Bolt 74B-A9B is designed to run agentic tasks locally, and SiliconANGLE reported that its two-bit quantization fits it within 24 gigabytes of memory.[2] NVIDIA said Super VL is five times faster than the next best model in its class, and that organizations can post-train it on their own data and run it on their own infrastructure.[1]