HF Model & Benchmark Matrix

24,078 variants · 14,984 scores · 120 benchmarks
Home/Orgs/nvidia/nvidia/NVIDIA-Nemotron-3-Nano-4B-BF16

nvidia/NVIDIA-Nemotron-3-Nano-4B-BF16

family nemotron · 7 variants · unknown

provider
nvidia
released
—
huggingface
https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Nano-4B-BF16
best scores
BFCL (function calling): 61.1IFBench: 44.2IFEval: 92.0RULER: 91.1AIME 2025: 78.5MATH-500: 95.4GPQA: 53.2LiveCodeBench: 51.8

Best scores per benchmark

benchmarkcategorybest score
MATH-500math95.4
IFEvalinstruction-following92.0
RULERlong-context91.1
AIME 2025math78.5
BFCL (function calling)agentic61.1
GPQAreasoning53.2
LiveCodeBenchcoding51.8
IFBenchinstruction-following44.2

Variants (7)

variantkindquantbitsvRAMdlorgyear
NVIDIA-Nemotron-3-Nano-4B-FP8fp8fp88.091,506nvidia2026
NVIDIA-Nemotron-3-Nano-4B-GGUFggufgguf4.046,371lmstudio-community2026
NVIDIA-Nemotron-3-Nano-4Boriginalbf1616.06,833unsloth2026
NVIDIA-Nemotron-3-Nano-4B-GGUFggufgguf16.05,749unsloth2026
nvidia_Nemotron-3-Nano-4B-GGUFggufgguf2.01,806bartowski2026
NVIDIA-Nemotron-3-Nano-4B-FP8fp8fp88.0332unsloth2026
Nemotron-3-Nano-4B-GGUFggufgguf16.0134ggml-org2026
Hugging Face model & benchmark data scraped & matrix-ified. 24,078 variants · 14,984 scores · 120 benchmarks.