Model
Model explorer

Nemotron-3 8B

OPEN
NVIDIA · Nemotron-3 family · released Nov 15, 2023

NVIDIA's first Nemotron-branded open LLM family, aimed at enterprise chat/copilot builds via NeMo.

ReasoningCodingVisionFunction callingTool useAgentic
39.5
Elo · rank #399
Parameters
8B
Active params
8B (dense)
Context
4K tokens
Architecture
Dense Transformer (GPT-3 style decoder-only)
License
NVIDIA AI Foundation Models Community License Agreement
Languages
53+
API price (in/out)
No hosted API
Modalities
text
Benchmark results
Bar shows position within the tracked field; marker = field best
ARC-EasyReasoning72.9%#41
best: Phi-3-medium (14B) · 97.7%
GSM8KMath39.4%#145
best: Llama 3.1 405B · 96.8%
HellaSwagReasoning76.4%#103
best: Claude 3 Opus · 95.4%
LAMBADAReasoning70.6%#17
best: GPT-3 175B · 86.4%
MMLUKnowledge54.4%#224
best: OpenAI o3 · 92.9%
PIQAReasoning80.4%#39
best: GPT-4o mini · 93.1%
WinoGrandeReasoning70.9%#97
best: PaLM 2 · 90.9%
Run it locally
VRAM @ Q4
5.4 GB
VRAM @ FP16
17 GB
Fits on (Q4)
RTX 3060 12GBRTX 4070 Ti 16GBRTX 3090 24GBRTX 4090 24GBRTX 5090 32GBM4 Pro 48GBM3 Max 128GBM3 Ultra 512GBA100 80GBH100 80GBH200 141GBB200 192GB
Throughput data unavailable.
Quantizations
GGUF Q4
Nemotron-3 family
Elo progression across releases
Nemotron-3 8BNov 202339.5
Resources
HFnvidia/nemotron-3-8b-base-4kDOCSdeveloper.nvidia.com
API price weights · each benchmark row carries its own source badge (see methodology)