← Back to all models

llama2 13b chat hf

Chat model by featherless-chat-models · 13.02B parameters

Chip
Unified Memory
Bandwidth 307 GB/s
Available for Models ~21 GB

Everything this chip runs, machine by machine: MacBook Pro 14" M5 Pro, MacBook Pro 16" M5 Pro

Runs Excellently

llama2 13b chat hf fits comfortably in your 24 GB Mac using Q8_0 quantization, using 63% of your RAM.

~12 tok/s

Q8_0 · 15.0 GB · Estimated

Quantization Options

QuantizationMemorySpeedFits?
Q8_0 Recommended 15.0 GB~12 tok/s✓
Q6_K 11.8 GB~16 tok/s✓
Q5_K_M 10.2 GB~19 tok/s✓
Q4_K_M 8.9 GB~22 tok/s✓
Q3_K_M 7.6 GB~27 tok/s✓
Q2_K 6.2 GB~35 tok/s✓

Specifications

Parameters13.0B
Architecturellama
Context Length4K
CategoryChat
Formatgguf
Minimum RAM7.3 GB
HuggingFace Downloads1,486
View on HuggingFace →

Which Mac Can Run llama2 13b chat hf?

Minimum Mac

MacBook Pro 14" — M1 Pro 16GB

Q6_K · ~11 tok/s · 74% RAM

MacQuanttok/sRAM %
Mac Studio · M3 Ultra 96GB Q8_0 ~33 16%
Mac Studio · M3 Ultra 256GB Q8_0 ~33 6%
Mac Studio · M3 Ultra 512GB Q8_0 ~33 3%
Mac Studio · M1 Ultra 64GB Q8_0 ~32 23%
Mac Studio · M1 Ultra 128GB Q8_0 ~32 12%
Mac Studio · M2 Ultra 64GB Q8_0 ~32 23%
Mac Studio · M2 Ultra 128GB Q8_0 ~32 12%
Mac Studio · M2 Ultra 192GB Q8_0 ~32 8%
MacBook Pro 14" · M5 Max 36GB Q8_0 ~25 42%
MacBook Pro 14" · M5 Max 48GB Q8_0 ~25 31%
MacBook Pro 14" · M5 Max 64GB Q8_0 ~25 23%
MacBook Pro 14" · M5 Max 128GB Q8_0 ~25 12%
MacBook Pro 16" · M5 Max 36GB Q8_0 ~25 42%
MacBook Pro 16" · M5 Max 48GB Q8_0 ~25 31%
MacBook Pro 16" · M5 Max 64GB Q8_0 ~25 23%
MacBook Pro 16" · M5 Max 128GB Q8_0 ~25 12%
MacBook Pro 14" · M4 Max 36GB Q8_0 ~22 42%
MacBook Pro 14" · M4 Max 48GB Q8_0 ~22 31%
MacBook Pro 14" · M4 Max 64GB Q8_0 ~22 23%
MacBook Pro 14" · M4 Max 128GB Q8_0 ~22 12%
MacBook Pro 16" · M4 Max 36GB Q8_0 ~22 42%
MacBook Pro 16" · M4 Max 48GB Q8_0 ~22 31%
MacBook Pro 16" · M4 Max 64GB Q8_0 ~22 23%
MacBook Pro 16" · M4 Max 128GB Q8_0 ~22 12%
Mac Studio · M4 Max 36GB Q8_0 ~22 42%
Mac Studio · M4 Max 48GB Q8_0 ~22 31%
Mac Studio · M4 Max 64GB Q8_0 ~22 23%
Mac Studio · M4 Max 128GB Q8_0 ~22 12%
MacBook Pro 14" · M1 Max 32GB Q8_0 ~16 47%
MacBook Pro 14" · M1 Max 64GB Q8_0 ~16 23%
MacBook Pro 14" · M2 Max 32GB Q8_0 ~16 47%
MacBook Pro 14" · M2 Max 64GB Q8_0 ~16 23%
MacBook Pro 14" · M2 Max 96GB Q8_0 ~16 16%
MacBook Pro 14" · M3 Max 36GB Q8_0 ~16 42%
MacBook Pro 14" · M3 Max 48GB Q8_0 ~16 31%
MacBook Pro 14" · M3 Max 64GB Q8_0 ~16 23%
MacBook Pro 14" · M3 Max 96GB Q8_0 ~16 16%
MacBook Pro 14" · M3 Max 128GB Q8_0 ~16 12%
MacBook Pro 16" · M1 Max 32GB Q8_0 ~16 47%
MacBook Pro 16" · M1 Max 64GB Q8_0 ~16 23%
MacBook Pro 16" · M2 Max 32GB Q8_0 ~16 47%
MacBook Pro 16" · M2 Max 64GB Q8_0 ~16 23%
MacBook Pro 16" · M2 Max 96GB Q8_0 ~16 16%
MacBook Pro 16" · M3 Max 36GB Q8_0 ~16 42%
MacBook Pro 16" · M3 Max 48GB Q8_0 ~16 31%
MacBook Pro 16" · M3 Max 64GB Q8_0 ~16 23%
MacBook Pro 16" · M3 Max 96GB Q8_0 ~16 16%
MacBook Pro 16" · M3 Max 128GB Q8_0 ~16 12%
Mac Studio · M1 Max 32GB Q8_0 ~16 47%
Mac Studio · M1 Max 64GB Q8_0 ~16 23%
Mac Studio · M2 Max 32GB Q8_0 ~16 47%
Mac Studio · M2 Max 64GB Q8_0 ~16 23%
Mac Studio · M2 Max 96GB Q8_0 ~16 16%
MacBook Pro 14" · M5 Pro 24GB Q8_0 ~12 63%
MacBook Pro 14" · M5 Pro 48GB Q8_0 ~12 31%
MacBook Pro 14" · M5 Pro 64GB Q8_0 ~12 23%
MacBook Pro 16" · M5 Pro 24GB Q8_0 ~12 63%
MacBook Pro 16" · M5 Pro 48GB Q8_0 ~12 31%
MacBook Pro 16" · M5 Pro 64GB Q8_0 ~12 23%
MacBook Pro 14" · M4 Pro 24GB Q8_0 ~11 63%
MacBook Pro 14" · M4 Pro 48GB Q8_0 ~11 31%
MacBook Pro 16" · M4 Pro 24GB Q8_0 ~11 63%
MacBook Pro 16" · M4 Pro 48GB Q8_0 ~11 31%
Mac mini · M4 Pro 24GB Q8_0 ~11 63%
Mac mini · M4 Pro 48GB Q8_0 ~11 31%
Mac mini · M4 Pro 64GB Q8_0 ~11 23%
MacBook Pro 14" · M1 Pro 16GB Q6_K ~11 74%
MacBook Pro 14" · M2 Pro 16GB Q6_K ~11 74%
MacBook Pro 16" · M1 Pro 16GB Q6_K ~11 74%
MacBook Pro 16" · M2 Pro 16GB Q6_K ~11 74%
Mac mini · M2 Pro 16GB Q6_K ~11 74%
MacBook Air · M5 16GB Q6_K ~8 74%
MacBook Pro 14" · M5 16GB Q6_K ~8 74%
MacBook Pro 14" · M1 Pro 32GB Q8_0 ~8 47%
MacBook Pro 14" · M2 Pro 32GB Q8_0 ~8 47%
MacBook Pro 16" · M1 Pro 32GB Q8_0 ~8 47%
MacBook Pro 16" · M2 Pro 32GB Q8_0 ~8 47%
Mac mini · M2 Pro 32GB Q8_0 ~8 47%
MacBook Air · M4 16GB Q6_K ~6 74%
MacBook Pro 14" · M4 16GB Q6_K ~6 74%
Mac mini · M4 16GB Q6_K ~6 74%
iMac · M4 16GB Q6_K ~6 74%
MacBook Air · M5 24GB Q8_0 ~6 63%
MacBook Air · M5 32GB Q8_0 ~6 47%
MacBook Pro 14" · M5 24GB Q8_0 ~6 63%
MacBook Pro 14" · M5 32GB Q8_0 ~6 47%
MacBook Pro 14" · M3 Pro 18GB Q8_0 ~6 83%
MacBook Pro 14" · M3 Pro 36GB Q8_0 ~6 42%
MacBook Pro 16" · M3 Pro 18GB Q8_0 ~6 83%
MacBook Pro 16" · M3 Pro 36GB Q8_0 ~6 42%
MacBook Air · M2 16GB Q6_K ~5 74%
MacBook Air · M3 16GB Q6_K ~5 74%
MacBook Pro 14" · M3 16GB Q6_K ~5 74%
Mac mini · M2 16GB Q6_K ~5 74%
iMac · M3 16GB Q6_K ~5 74%
MacBook Air · M4 24GB Q8_0 ~5 63%
MacBook Air · M4 32GB Q8_0 ~5 47%
MacBook Pro 14" · M4 24GB Q8_0 ~5 63%
MacBook Pro 14" · M4 32GB Q8_0 ~5 47%
Mac mini · M4 24GB Q8_0 ~5 63%
Mac mini · M4 32GB Q8_0 ~5 47%
iMac · M4 24GB Q8_0 ~5 63%
iMac · M4 32GB Q8_0 ~5 47%
MacBook Air · M2 24GB Q8_0 ~4 63%
MacBook Air · M3 24GB Q8_0 ~4 63%
MacBook Pro 14" · M3 24GB Q8_0 ~4 63%
Mac mini · M2 24GB Q8_0 ~4 63%
iMac · M3 24GB Q8_0 ~4 63%
MacBook Air · M1 16GB Q6_K ~4 74%
Mac mini · M1 16GB Q6_K ~4 74%
iMac · M1 16GB Q6_K ~4 74%
MacBook Air · M2 8GB Q2_K ~11 77%
MacBook Air · M3 8GB Q2_K ~11 77%
MacBook Pro 14" · M3 8GB Q2_K ~11 77%
Mac mini · M2 8GB Q2_K ~11 77%
iMac · M3 8GB Q2_K ~11 77%
MacBook Air · M1 8GB Q2_K ~8 77%
Mac mini · M1 8GB Q2_K ~8 77%
iMac · M1 8GB Q2_K ~8 77%

Models similar to llama2 13b chat hf

Closest matches by use case and parameter count.

Run llama2 13b chat hf locally on your Mac

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

Get ToolPiper — Free