NVIDIA’s 550B Nemotron Embarrassed Every US Open Model — and It Shouldn’t Run This Fast
Last Updated on June 3, 2026 by Editorial Team
Author(s): Chew Loong Nian – AI ENGINEER
Originally published on Towards AI.
NVIDIA's 550B Nemotron Embarrassed Every US Open Model — and It Shouldn't Run This Fast
NVIDIA just shipped a 550B-parameter open model that scores 48 on the Artificial Analysis Intelligence Index. The next-best American open-weights model, Google’s Gemma 4, sits at 39. OpenAI’s gpt-oss-120b sits at 33. NVIDIA’s own previous flagship, Nemotron 3 Super, sits at 36.
Summary of the article: After detailing NVIDIA’s Computex announcement of Nemotron 3 Ultra (about 500B–550B parameters, released across major model hubs, and evaluated alongside partners), the article explains the hybrid Mamba/mixture-of-experts architecture designed to make a huge model run efficiently by activating far fewer parameters per token. It compares benchmarks, noting Ultra matches the Chinese frontier on areas like agentic/search performance and long-context accuracy at 1M tokens, but still trails the top Chinese model (Kimi K2.6) on raw intelligence scores. The author reframes the “real win” as throughput and cost—Ultra’s much higher tokens-per-second and cheaper serving make it more practical for deploying agents—then outlines where users can get the models now (Ultra via hosted routes, Nano/Super for local use) and closes with a verdict emphasizing efficiency as the metric that matters for most teams.
Read the full blog for free on Medium.
Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming a sponsor.
Published via Towards AI
Towards AI Academy
We Build Enterprise-Grade AI. We'll Teach You to Master It Too.
15 engineers. 100,000+ students. Towards AI Academy teaches what actually survives production.
Start free — no commitment:
→ 6-Day Agentic AI Engineering Email Guide — one practical lesson per day
→ Agents Architecture Cheatsheet — 3 years of architecture decisions in 6 pages
Our courses:
→ AI Engineering Certification — 90+ lessons from project selection to deployed product. The most comprehensive practical LLM course out there.
→ Agent Engineering Course — Hands on with production agent architectures, memory, routing, and eval frameworks — built from real enterprise engagements.
→ AI for Work — Understand, evaluate, and apply AI for complex work tasks.
Note: Article content contains the views of the contributing authors and not Towards AI.
Related posts
Recent Posts
