← Back to all models

Llama 3.2 3B Instruct FP8

Chat model by redhatai · 3.61B parameters

Runs Excellently

Llama 3.2 3B Instruct FP8 fits comfortably in your 24 GB Mac using Q8_0 quantization, using 19% of your RAM.

~40 tok/s

Q8_0 · 4.5 GB

Quantization Options

QuantizationMemorySpeedFits?
Q8_0 Recommended 4.5 GB~40 tok/s
Q6_K 3.6 GB~52 tok/s
Q5_K_M 3.2 GB~61 tok/s
Q4_K_M 2.8 GB~72 tok/s
Q3_K_M 2.5 GB~87 tok/s
Q2_K 2.1 GB~112 tok/s

Specifications

Parameters3.6B
Architecturellama
Context Length131K
CategoryChat
Capabilitiestool_use
Formatgguf
Minimum RAM2 GB
HuggingFace Downloads59,891

Which Mac Can Run Llama 3.2 3B Instruct FP8?

Minimum Mac

MacBook Air — M1 8GB

Q8_0 · ~10 tok/s · 57% RAM

MacQuanttok/sRAM %
MacBook Air · M1 8GB Q8_0 ~10 57%
MacBook Air · M2 8GB Q8_0 ~15 57%
MacBook Air · M3 8GB Q8_0 ~15 57%
Mac Mini · M1 8GB Q8_0 ~10 57%
Mac Mini · M2 8GB Q8_0 ~15 57%
iMac · M1 8GB Q8_0 ~10 57%
iMac · M3 8GB Q8_0 ~15 57%
MacBook Air · M1 16GB Q8_0 ~10 28%
MacBook Air · M2 16GB Q8_0 ~15 28%
MacBook Air · M3 16GB Q8_0 ~15 28%
MacBook Air · M4 16GB Q8_0 ~17 28%
MacBook Air · M5 16GB Q8_0 ~22 28%
MacBook Air · M6 16GB Q8_0 ~30 28%
MacBook Pro 14" · M1 Pro 16GB Q8_0 ~29 28%
MacBook Pro 14" · M2 Pro 16GB Q8_0 ~29 28%
MacBook Pro 14" · M5 16GB Q8_0 ~22 28%
MacBook Pro 14" · M6 16GB Q8_0 ~30 28%
MacBook Pro 16" · M1 Pro 16GB Q8_0 ~29 28%
MacBook Pro 16" · M2 Pro 16GB Q8_0 ~29 28%
Mac Mini · M1 16GB Q8_0 ~10 28%
Mac Mini · M2 16GB Q8_0 ~15 28%
Mac Mini · M2 Pro 16GB Q8_0 ~29 28%
Mac Mini · M4 16GB Q8_0 ~17 28%
Mac Mini · M6 16GB Q8_0 ~30 28%
iMac · M1 16GB Q8_0 ~10 28%
iMac · M3 16GB Q8_0 ~15 28%
iMac · M4 16GB Q8_0 ~17 28%
iMac · M6 16GB Q8_0 ~30 28%
MacBook Pro 14" · M3 Pro 18GB Q8_0 ~22 25%
MacBook Pro 16" · M3 Pro 18GB Q8_0 ~22 25%
MacBook Air · M2 24GB Q8_0 ~15 19%
MacBook Air · M3 24GB Q8_0 ~15 19%
MacBook Air · M4 24GB Q8_0 ~17 19%
MacBook Air · M5 24GB Q8_0 ~22 19%
MacBook Air · M6 24GB Q8_0 ~30 19%
MacBook Pro 14" · M4 Pro 24GB Q8_0 ~40 19%
MacBook Pro 14" · M5 24GB Q8_0 ~22 19%
MacBook Pro 14" · M5 Pro 24GB Q8_0 ~45 19%
MacBook Pro 14" · M6 24GB Q8_0 ~30 19%
MacBook Pro 14" · M6 Pro 24GB Q8_0 ~59 19%
MacBook Pro 16" · M4 Pro 24GB Q8_0 ~40 19%
MacBook Pro 16" · M5 Pro 24GB Q8_0 ~45 19%
MacBook Pro 16" · M6 Pro 24GB Q8_0 ~59 19%
Mac Mini · M2 24GB Q8_0 ~15 19%
Mac Mini · M4 24GB Q8_0 ~17 19%
Mac Mini · M4 Pro 24GB Q8_0 ~40 19%
Mac Mini · M6 24GB Q8_0 ~30 19%
Mac Mini · M6 Pro 24GB Q8_0 ~59 19%
iMac · M3 24GB Q8_0 ~15 19%
iMac · M4 24GB Q8_0 ~17 19%
iMac · M6 24GB Q8_0 ~30 19%
MacBook Air · M4 32GB Q8_0 ~17 14%
MacBook Air · M5 32GB Q8_0 ~22 14%
MacBook Air · M6 32GB Q8_0 ~30 14%
MacBook Pro 14" · M1 Pro 32GB Q8_0 ~29 14%
MacBook Pro 14" · M2 Pro 32GB Q8_0 ~29 14%
MacBook Pro 14" · M5 32GB Q8_0 ~22 14%
MacBook Pro 14" · M6 32GB Q8_0 ~30 14%
MacBook Pro 16" · M1 Pro 32GB Q8_0 ~29 14%
MacBook Pro 16" · M1 Max 32GB Q8_0 ~58 14%
MacBook Pro 16" · M2 Pro 32GB Q8_0 ~29 14%
MacBook Pro 16" · M2 Max 32GB Q8_0 ~58 14%
Mac Mini · M2 Pro 32GB Q8_0 ~29 14%
Mac Mini · M4 32GB Q8_0 ~17 14%
Mac Mini · M6 32GB Q8_0 ~30 14%
Mac Studio · M1 Max 32GB Q8_0 ~58 14%
Mac Studio · M2 Max 32GB Q8_0 ~58 14%
iMac · M4 32GB Q8_0 ~17 14%
iMac · M6 32GB Q8_0 ~30 14%
MacBook Pro 14" · M3 Pro 36GB Q8_0 ~22 13%
MacBook Pro 14" · M5 Max 36GB Q8_0 ~89 13%
MacBook Pro 16" · M3 Pro 36GB Q8_0 ~22 13%
MacBook Pro 16" · M3 Max 36GB Q8_0 ~58 13%
MacBook Pro 16" · M4 Max 36GB Q8_0 ~79 13%
MacBook Pro 16" · M5 Max 36GB Q8_0 ~89 13%
Mac Studio · M4 Max 36GB Q8_0 ~79 13%
MacBook Pro 14" · M4 Pro 48GB Q8_0 ~40 9%
MacBook Pro 14" · M5 Pro 48GB Q8_0 ~45 9%
MacBook Pro 14" · M5 Max 48GB Q8_0 ~89 9%
MacBook Pro 14" · M6 Pro 48GB Q8_0 ~59 9%
MacBook Pro 14" · M6 Max 48GB Q8_0 ~134 9%
MacBook Pro 16" · M3 Max 48GB Q8_0 ~58 9%
MacBook Pro 16" · M4 Pro 48GB Q8_0 ~40 9%
MacBook Pro 16" · M4 Max 48GB Q8_0 ~79 9%
MacBook Pro 16" · M5 Pro 48GB Q8_0 ~45 9%
MacBook Pro 16" · M5 Max 48GB Q8_0 ~89 9%
MacBook Pro 16" · M6 Pro 48GB Q8_0 ~59 9%
MacBook Pro 16" · M6 Max 48GB Q8_0 ~134 9%
Mac Mini · M4 Pro 48GB Q8_0 ~40 9%
Mac Mini · M6 Pro 48GB Q8_0 ~59 9%
Mac Studio · M6 Max 48GB Q8_0 ~134 9%
MacBook Pro 14" · M5 Pro 64GB Q8_0 ~45 7%
MacBook Pro 14" · M5 Max 64GB Q8_0 ~89 7%
MacBook Pro 14" · M6 Pro 64GB Q8_0 ~59 7%
MacBook Pro 14" · M6 Max 64GB Q8_0 ~134 7%
MacBook Pro 16" · M1 Max 64GB Q8_0 ~58 7%
MacBook Pro 16" · M2 Max 64GB Q8_0 ~58 7%
MacBook Pro 16" · M3 Max 64GB Q8_0 ~58 7%
MacBook Pro 16" · M4 Max 64GB Q8_0 ~79 7%
MacBook Pro 16" · M5 Max 64GB Q8_0 ~89 7%
MacBook Pro 16" · M6 Pro 64GB Q8_0 ~59 7%
MacBook Pro 16" · M6 Max 64GB Q8_0 ~134 7%
Mac Mini · M6 Pro 64GB Q8_0 ~59 7%
Mac Studio · M1 Max 64GB Q8_0 ~58 7%
Mac Studio · M1 Ultra 64GB Q8_0 ~116 7%
Mac Studio · M2 Max 64GB Q8_0 ~58 7%
Mac Studio · M2 Ultra 64GB Q8_0 ~116 7%
Mac Studio · M4 Max 64GB Q8_0 ~79 7%
Mac Studio · M6 Max 64GB Q8_0 ~134 7%
MacBook Pro 16" · M2 Max 96GB Q8_0 ~58 5%
MacBook Pro 16" · M3 Max 96GB Q8_0 ~58 5%
Mac Studio · M2 Max 96GB Q8_0 ~58 5%
Mac Studio · M2 Ultra 96GB Q8_0 ~116 5%
Mac Pro · M2 Ultra 96GB Q8_0 ~116 5%
MacBook Pro 14" · M5 Max 128GB Q8_0 ~89 4%
MacBook Pro 14" · M6 Max 128GB Q8_0 ~134 4%
MacBook Pro 16" · M3 Max 128GB Q8_0 ~58 4%
MacBook Pro 16" · M4 Max 128GB Q8_0 ~79 4%
MacBook Pro 16" · M5 Max 128GB Q8_0 ~89 4%
MacBook Pro 16" · M6 Max 128GB Q8_0 ~134 4%
Mac Studio · M1 Ultra 128GB Q8_0 ~116 4%
Mac Studio · M2 Ultra 128GB Q8_0 ~116 4%
Mac Studio · M4 Max 128GB Q8_0 ~79 4%
Mac Studio · M4 Ultra 128GB Q8_0 ~119 4%
Mac Studio · M6 Max 128GB Q8_0 ~134 4%
Mac Pro · M2 Ultra 128GB Q8_0 ~116 4%
Mac Pro · M4 Ultra 128GB Q8_0 ~119 4%
MacBook Pro 14" · M6 Max 192GB Q8_0 ~134 2%
MacBook Pro 16" · M6 Max 192GB Q8_0 ~134 2%
MacBook Pro 16" · M6 Ultra 192GB Q8_0 ~267 2%
Mac Studio · M2 Ultra 192GB Q8_0 ~116 2%
Mac Studio · M4 Ultra 192GB Q8_0 ~119 2%
Mac Studio · M6 Max 192GB Q8_0 ~134 2%
Mac Studio · M6 Ultra 192GB Q8_0 ~267 2%
Mac Pro · M2 Ultra 192GB Q8_0 ~116 2%
Mac Pro · M4 Ultra 192GB Q8_0 ~119 2%
Mac Pro · M6 Ultra 192GB Q8_0 ~267 2%
MacBook Pro 16" · M6 Ultra 256GB Q8_0 ~267 2%
Mac Studio · M4 Ultra 256GB Q8_0 ~119 2%
Mac Studio · M6 Ultra 256GB Q8_0 ~267 2%
Mac Pro · M4 Ultra 256GB Q8_0 ~119 2%
Mac Pro · M6 Ultra 256GB Q8_0 ~267 2%
MacBook Pro 16" · M6 Ultra 384GB Q8_0 ~267 1%
Mac Studio · M6 Ultra 384GB Q8_0 ~267 1%
Mac Pro · M6 Ultra 384GB Q8_0 ~267 1%

Run Llama 3.2 3B Instruct FP8 locally on your Mac

ToolPiper downloads, manages, and runs models with one click. Apple Silicon optimized.

Get ToolPiper — Free