Compare AI models

Specifications
Parameters
550B4B/30B/235B
Context window
1M256k
API pricing
Cheapest input
$0.40DeepInfra—
Cheapest output
$1.76DeepInfra—
$210KT—
Benchmarks
LiveCodeBench
89%74.1%
Humanity's Last Exam · text only, no tools
29.24%18.2%
GPQA Diamond
87%81.1%
MMLU-Pro
86.8%84.4%
Overview
CompanyNVIDIAQwen
Release dateJun 4 2026Jul 21 2025
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

Nemotron 3 Ultra leads Qwen3-2507 on 4 of the 4 benchmarks they both report (LiveCodeBench, Humanity's Last Exam, GPQA Diamond, MMLU-Pro). Qwen3-2507 shipped 318 days before Nemotron 3 Ultra, so benchmark comparisons should account for the intervening progress.

Nemotron 3 Ultra has 550B parameters, while Qwen3-2507 has 4B/30B/235B. Context windows are 1M (Nemotron 3 Ultra) vs 256k (Qwen3-2507).

On LiveCodeBench, Nemotron 3 Ultra leads at 89% vs Qwen3-2507 at 74.1%. On Humanity's Last Exam · text only, no tools, Nemotron 3 Ultra leads at 29.24% vs Qwen3-2507 at 18.2%. On GPQA Diamond, Nemotron 3 Ultra leads at 87% vs Qwen3-2507 at 81.1%. On MMLU-Pro, Nemotron 3 Ultra leads at 86.8% vs Qwen3-2507 at 84.4%.