Perks
NVIDIA ships Nemotron 3.5 Super VL, an open AI model for agents, plus two Bolt models — image from NVIDIA
Image: NVIDIA

BriefsNewsAvailable now1 min read

NVIDIA ships Nemotron 3.5 Super VL, an open AI model for agents, plus two Bolt models

The open 120-billion-parameter Super VL and the 24GB Bolt 74B-A9B are out now. A smaller Bolt E4B follows on Oct. 23.

Summary

NVIDIA released Nemotron 3.5 Super VL, an open 120-billion-parameter model for always-on agents. It understands images, webpages and PDFs, then acts on them. Two smaller Bolt models run locally: Bolt 74B-A9B is out now, and Bolt E4B arrives Oct. 23.

What changed

  • Super VL is on Hugging Face, Open Router and the ModelScope community, and also comes as NVIDIA NIM.
  • Bolt 74B-A9B 24GB uses native 2-bit quantization to fit in 24GB of memory. It runs everyday agent tasks on NVIDIA RTX/DGX Spark.
  • Bolt E4B 4GB targets fast conversational agents, such as voice assistants and embedded interfaces. It arrives Oct. 23.
  • Super VL comes with open weights, open datasets and training recipes.

Availability

Available now

“Super VL and Bolt 74B-A9B are available now, with Bolt E4B coming Oct. 23.”

Founder takeaway

If you build agents that read documents or webpages, or want one running on your own RTX hardware, try Super VL or Bolt 74B-A9B now. Bolt E4B isn't out until Oct. 23, so there's nothing to do with it yet.

Source

NVIDIAOpen From Data Center to Desktop: NVIDIA Expands Nemotron Family With Highest Efficiency Super VL and Bolt Models

Announced Checked

Researched and drafted with AI from the official sources above, then checked automatically before publishing. Spotted a mistake? Tell us.

SharePost on XLinkedIn

Related

More briefs

All briefs
  1. Google shows off the Gemini agent, an AI assistant and teammate in one — image from Google Cloud

    NewsAnnounced — not available yet

    Google shows off the Gemini agent, an AI assistant and teammate in one

    Google's new Gemini agent AI is meant to be both a personal assistant and a team member. It isn't public yet, but startups can talk to Google's sales team.

    Google Cloud1 min read

  2. GitHub's AI classifier for secret scanning cuts false alerts and scan time — image from GitHub

    NewsAvailable now

    GitHub's AI classifier for secret scanning cuts false alerts and scan time

    GitHub describes a fine-tuned ModernBERT model that sorts real credentials from lookalikes, replacing an LLM-based pipeline in secret scanning.

    GitHub1 min read

  3. Vercel adds NVIDIA Nemotron 3.5 Lightning to AI Gateway for free — image from Vercel

    NewsAvailable now

    Vercel adds NVIDIA Nemotron 3.5 Lightning to AI Gateway for free

    Vercel's AI Gateway now offers NVIDIA's open-weight Nemotron 3.5 Lightning model at no charge, with a note that the provider may keep prompts.

    Vercel1 min read

One email, every Monday

Briefs are published here every weekday. Monday Heist is the weekly email: one useful AI workflow, the files to use it, and the catch. Free, and you can leave in one click.

One email. Every Monday.