Nvidia announces Nemotron 3.5 Super VL and smaller Bolt models for agent workloads

Nvidia announced Nemotron 3.5 Super VL, a 120-billion-parameter vision-language model for enterprise agents, alongside two smaller Nemotron Bolt models. Super VL combines reasoning, tool use and image and document understanding for multimodal workflows. The Bolt lineup includes a 74B-A9B model designed to fit in 24GB and an E4B model for 4GB, targeting local coding, research and conversational tasks. The two larger models are available now; the smaller E4B model is planned for a later launch.

AI-written summary, checked against the cited sources. How summaries work

Image: NVIDIA Developer Blog

Why it matters

Teams building local agent workloads can evaluate the two larger models now; the Nemotron 3.5 Bolt E4B model is expected to launch later.

Release

Version
3.5
Released

What we know

Confirmed by several sources

Reported by one source

  • The Super VL model uses a mixture-of-experts architecture with 12 billion active parameters. — NVIDIA Developer Blog
  • Nvidia announced two Bolt models: a 74B-A9B model sized for 24GB and an E4B model sized for 4GB. — The New Stack
  • The two larger models are available today, while the E4B model is planned for a future launch. — The New Stack
  • Nvidia says Super VL outperforms frontier models on specialized benchmarks. — Techmeme

Coverage 3 publishers

  1. NVIDIA Developer BlogFirst reportPrimary source

    NVIDIA Nemotron 3.5 Super VL Powers Specialized Agents with Multimodal Reasoning

    Always-on agents are taking on longer, more complex enterprise workflows. A single request can involve reading a document, checking information across sources,...Always-on agents are taking on longer, more complex enterprise workflows. A…

  2. The New Stack

    Nvidia squeezed a 74B agent model into 24GB so it can run on your laptop

    Nvidia on Thursday announced a few new additions to its Nemotron family: Nemotron 3.5 Super VL, a 120B vision-language model for enterprise agents and two small Bolt models (Nemotron 3.5 Bolt 74B-A9B 24GB and Nemotron 3.5 Bolt E4B 4GB) for…

  3. Techmeme

    Nvidia launches Nemotron 3.5 Super VL, a 120B-parameter open model for enterprise agents, and says it outperforms frontier models on specialized benchmarks

    An enterprise agent reviewing a credit agreement may need to extract figures from tables, interpret a chart, check a related webpage …

Earliest report first. Articles stay on their publishers’ sites; each link opens the original.