Saltar al contenido principal

Nemotron 3 Super 120B A12B (Reasoning)

NVIDIAOpen WeightNVIDIA Open Model License Agreement · Uso Comercial

Descripción

Nemotron 3 Super is a 120B total / 12B active parameter hybrid Mamba-Attention Mixture-of-Experts model optimized for agentic reasoning, coding, planning, tool calling, and long-context analysis. It introduces LatentMoE (projecting tokens into a compressed latent space for expert routing, enabling 4x more experts at the same inference cost), Multi-Token Prediction for native speculative decoding (up to 3x faster generation), and native NVFP4 pretraining on Blackwell. The hybrid architecture interleaves Mamba-2 layers for linear-time sequence processing with strategically placed Transformer attention layers as global anchors, supporting a 1M-token context window. Pre-trained on 25 trillion tokens and post-trained with multi-environment RL across 21 configurations using NeMo Gym/RL with 1.2 million rollouts. Achieves up to 5x higher throughput than previous Nemotron Super and 2.2x higher throughput than GPT-OSS-120B while maintaining comparable accuracy.

Fecha de lanzamiento
2026-03-11
Parámetros
120.0B
Longitud del contexto
262K
Modalidades
text

Radar de capacidades

60
general
50
coding
70
reasoning
51
scienceest.
50
agents
0
multimodal

Science se estima a partir de las puntuaciones científicas de LLM Stats o del razonamiento cuando no hay benchmarks científicos dedicados.

Rankings

No hay datos de ranking disponibles

Puntuaciones de benchmarks (LLM Stats)

(LLM Stats (zeroeval))

Chat

Tau2 Airline56.3%Aut.
Multi-Challenge55.2%Aut.

Communication

Tau2 Telecom64.4%Aut.
Tau2 Retail62.8%Aut.

General

Arena-Hard v273.9%Aut.

Instruction Following

IFBench72.6%Aut.

Language

WMT24++86.7%Aut.
MMLU-Pro83.7%Aut.
MMLU-ProX79.4%Aut.

Long Context

RULER91.8%Aut.

Math

HMMT 202594.7%Aut.
AIME 202590.2%Aut.

Reasoning

GPQANYU + Cohere + Anthropic (2023)82.7%Aut.
LiveCodeBench81.2%Aut.
AA-LCR58.3%Aut.
SWE-Bench Verified53.7%Aut.
SWE-bench Multilingual45.8%Aut.
SciCode42.0%Aut.
Bird-SQL (dev)41.8%Aut.
BrowseCompOpenAI (2025)31.3%Aut.
Terminal-Bench 2.0Stanford × Laude Institute (2026)31.0%Aut.
Terminal-Bench25.8%Aut.
Humanity's Last Exam22.8%Aut.

Índices de evaluación AA

(Artificial Analysis)

No hay datos de evaluación AA disponibles

Puntuaciones por categoría LLM Stats

(LLM Stats (zeroeval))
Language
80
Legal
80
Long Context
80
Finance
80
Healthcare
80
Instruction Following
70
Math
70
Creativity
70
Writing
70
Chat
60
Physics
60
Reasoning
60
General
60
Biology
60
Chemistry
60
Communication
60
Frontend Development
50
Code
50
Tool Calling
50
Search
30
Agents
30
Vision
20

Precios

Precio de entrada$0.3 / 1M tokens
Precio de salida$0.9 / 1M tokens
Precio mixto (3:1)$0.16375 / 1M tokens

Velocidad

Tokens/seg193.7
Retraso del primer token1.06s
Tiempo hasta la respuesta11.38s

Ranking de Precios por Proveedor

Ranking de Precios por Proveedor

9 proveedores

Más barato: NanoGPTMás caro: Nebius Token Factory
ProveedorEntradaSalida
1NanoGPTMás barato
$0.05
$0.25
2Kilo Gateway
$0.06
$0.32
3OpenRouter
$0.075
$0.4
4DeepInfra
$0.085
$0.4
5Requesty
$0.1
$0.5
6Vercel AI Gateway
$0.15
$0.65
7Perplexity Agent
$0.25
$2.5
8NVIDIAPRINCIPAL
$0.3
$0.9
9Nebius Token Factory
$0.3
$0.9

Comparar precios entre diferentes proveedores de API para este modelo.

Fuentes externas