Skip to content

NVIDIA Nemotron 3 Super 120B A12B (Reasoning)

Available

Other · 2026-03-11 · 32,000 tokens

An AI model from Other, suited to a broad range of AI workloads.

Supported modalities:textcode

Quick Overview

Text Generation3/10
Code Generation4/10
Reasoning6/10
Multimodal2/10

Benchmark Results

Scores from leading benchmark suites.

artificial analysis intelligence25.4
artificial analysis coding37.7

Performance Metrics

Latency and throughput performance.

1000msP50 Latency
142.982tokens/sec

Dive Deeper