HF Model & Benchmark Matrix

24,078 variants · 14,984 scores · 120 benchmarks
Home/Orgs/meta-llama/meta-llama/Llama-3.2-3B

meta-llama/Llama-3.2-3B

family llama · 11 variants · unknown

provider
meta-llama
released
—
huggingface
https://huggingface.co/meta-llama/Llama-3.2-3B
best scores
MATH: 44.4MATH-500: 44.4GPQA Diamond: 27.78

Best scores per benchmark

benchmarkcategorybest score
MATHmath44.4
MATH-500math44.4
GPQA Diamondreasoning27.78

Variants (11)

variantkindquantbitsvRAMdlorgyear
Llama-3.2-3Boriginalbf1616.02,529535,893meta-llama2024
Llama-3.2-3Boriginalbf1616.047,917unsloth2025
Llama-3.2-3B-bnb-4bitbnb-4bit4bit11,032unsloth2025
tada-3b-mloriginalbf1616.07,992HumeAI2026
llama-nv-embed-reasoning-3boriginalbf1616.02,141nvidia2026
Dolphin3.0-Llama3.2-3B-GGUFggufgguf16.0411Mungert2025
SIMS-Llama3.2-3Boriginalbf1616.0395slprl2025
Llama-3.2-3B-Instruct-GGUFggufgguf16.0219Mungert2025
Llama-3.2-3B-F1-Reasoning-Instruct-GGUFggufgguf16.0122Mungert2025
Llama-3.2-3B-Instruct-QLORA_INT4_EO8originalbf1616.078meta-llama2024
Llama-3.2-3B-Instruct-SpinQuant_INT4_EO8originalbf1616.039meta-llama2024
Hugging Face model & benchmark data scraped & matrix-ified. 24,078 variants · 14,984 scores · 120 benchmarks.