Compare AI models

Specifications
Parameters
550B14.3B
Context window
1M32k
API pricing
Cheapest input
$0.40DeepInfra—
Cheapest output
$1.76DeepInfra—
$210KT—
These models have no shared benchmark scores.
Overview
CompanyNVIDIAQwen
Release dateJun 4 2026Mar 28 2024
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

Nemotron 3 Ultra and Qwen1.5-MoE don't publish scores on any of the same benchmarks, so there's no direct head-to-head comparison. Qwen1.5-MoE shipped 798 days before Nemotron 3 Ultra, so benchmark comparisons should account for the intervening progress.

Nemotron 3 Ultra has 550B parameters, while Qwen1.5-MoE has 14.3B. Context windows are 1M (Nemotron 3 Ultra) vs 32k (Qwen1.5-MoE).

Direct benchmark comparisons are unavailable — Nemotron 3 Ultra and Qwen1.5-MoE don't publish scores on any of the same benchmarks.