← Back
NVIDIA Blog

NVIDIA Open Models Advance Europe’s Sovereign Intelligence Across Industries

•6 min read•
#llm#agents#deployment
Level:Intermediate
For:AI Engineers
✦TL;DR

NVIDIA announced an expanded portfolio of open‑weight models—including Nemotron 3.5 Super VL, Nemotron 3.5 Bolt, Nemotron Retriever, Kumo Structured, Kumo Engine, Cosmos 3.5, Alpamayo 2 Edge, BioNeMo, and Ising—to let developers and enterprises build language, robotics, biology, and quantum‑computing applications tailored to local languages and industries. The Nemotron models support text and images for agent workflows, while Kumo provides pretrained structured‑data foundations and an open library stack for rapid prediction pipelines. These models are open‑weight, can be fine‑tuned with proprietary data, and are intended for deployment on self‑controlled infrastructure. Engineers can thus accelerate domain‑specific AI development without building separate ML pipelines from scratch.

⚡ Key Takeaways

  • New open‑weight models released: Nemotron 3.5 Super VL, Nemotron 3.5 Bolt, Nemotron Retriever, Kumo Structured, Kumo Engine, Cosmos 3.5, Alpamayo 2 Edge, BioNeMo, Ising.
  • Nemotron models support both text and image inputs for multi‑modal agent orchestration.
  • Kumo Engine is an open library stack for developing and deploying structured‑data foundation models, reducing prediction latency and cost.
  • Fine‑tuning is encouraged: developers can start with pretrained Kumo or Nemotron models and adapt them with proprietary datasets for higher performance.
  • Deployment is on infrastructure the organization controls; no cloud‑only lock‑in is required.
💡 Why It Matters

By providing open‑weight, multi‑modal foundations and a ready‑to‑use structured‑data stack, NVIDIA lets enterprises rapidly prototype and productionize domain‑specific AI services while keeping data and compute in‑house, lowering both development time and operational cost.

✅ Practical Steps

  1. Import the Kumo Engine library stack and load a pretrained Kumo Structured model to add tabular prediction capabilities to your application.
  2. Fine‑tune a Nemotron 3.5 Bolt model on your proprietary text‑image dataset to create a domain‑specific agent that can reason over specialized documents.

Want the full story? Read the original article.

Read on NVIDIA Blog ↗

More like this

Censored vs. Truncated Regression in Python, Explained

Towards Data Science•#llm

New in AWS Quick: More capable flows, an updated web experience, and agents that work together

AWS ML Blog•#agents

Three insights you may have missed from theCUBE’s coverage of the Dell Technologies New York Symposium

SiliconANGLE AI•#deployment

How Google Duplex evolved to also call businesses for you in Gemini

Google Research•#agents

EXPLORE AI NEWS

Daily hand-picked stories on LLMs, RAG, agents and production AI — curated for engineers who ship.

BROWSE NEWS

GET THE WEEKLY DIGEST

Join engineers getting the Monday signal-over-noise AI breakdown. No spam, unsubscribe anytime.

LEARN AI ENGINEERING

Curated courses, research papers, repos and tutorials built for engineers leveling up in AI.

START LEARNING