Models/NVIDIA-Nemotron-Nano-9B-v2

MDL

NVIDIA-Nemotron-Nano-9B-v2

nvidia

Hugging Face ↗

Verified against Artificial Analysis · Jul 13, 2026

Parameters

Price /M

$0.04 / $0.16

in / out

Context

—

Avg score

37.2

10 benchmarks

AA Intelligence

8.8

index

Speed

79 tok/s

TTFT 7s

Released

Aug 18, 2025

About

NVIDIA-Nemotron-Nano-9B-v2 is a 9B-parameter open-weights model in the NVIDIA family from nvidia. Benchmarked on 10 evals, averaging 37.2.

othertext-generation

Benchmark scores · 10

avg 37.2 — higher is better

MMLU-Proartificial-analysis · 2026-07

74.2

LiveCodeBenchartificial-analysis · 2026-07

72.4

AIMEartificial-analysis · 2026-07

69.7

GPQA Diamondartificial-analysis · 2026-07

IFBenchartificial-analysis · 2026-07

27.6

SciCodeartificial-analysis · 2026-07

τ²-benchartificial-analysis · 2026-07

21.9

AA Long-Context Reasoningartificial-analysis · 2026-07

Humanity's Last Examartificial-analysis · 2026-07

4.6

Terminal-Bench Hardartificial-analysis · 2026-07

1.5

The research behind this model

all papers mentioning it →

Nemotron-H: A Family of Accurate and Efficient Hybrid Mamba-Transformer Models NVIDIA Nemotron Nano 2: An Accurate and Efficient Hybrid Mamba-Transformer Reasoning Model Nemotron-CC: Transforming Common Crawl into a Refined Long-Horizon Pretraining Dataset

Details

Publishernvidia

FamilyNVIDIA

Parameters9B

Licenseother

WeightsOpen

ReleasedAug 18, 2025

Performance & price

Output speed79 tok/s

Latency (TTFT)7s

First answer32s

Price in / out$0.04 / $0.16 per 1M

Blended price$0.07 / 1M

Math index69.7

Median across API providers, via Artificial Analysis. Blended = 3:1 input:output. For reasoning models, latency includes thinking time. Methodology →

History

full ledger →

Capability over time

Append-only ledger — every observed change to this model's numbers.

API providers

Artificial Analysis$0.16/M