onmydevice
.ai
⌘
⌘
Devices
Models
Benchmarks
Apps
Replace
Blog
Llama 3.2 3B
2.0 GB
Chat
128K context
·
Requires Mid+ tier device
Meta's 3B — best small model for many tasks
MIN TIER
Mid
SMALLEST QUANT
Q2 · 1.1 GB
CONTEXT WINDOW
128K tokens
PARAMETERS
3.21B
QUANTIZATION OPTIONS
Quant
File size
Quality
FP16
6.6 GB
100%
Best quality
Q8
3.6 GB
95%
Q6
2.6 GB
85%
Q5
2.3 GB
78%
Q4
1.9 GB
70%
Q3
1.4 GB
58%
Q2
1.1 GB
42%
DEVICE COMPATIBILITY
Runs on these devices (47)
💻
MacBook Pro M1 Max
32GB
macOS
FP16
·
~61 tok/s
·
6.6 GB
Runs great
💻
MacBook Pro M1 Max
64GB
macOS
FP16
·
~61 tok/s
·
6.6 GB
Runs great
💻
MacBook Pro M2 Max
32GB
macOS
FP16
·
~61 tok/s
·
6.6 GB
Runs great
💻
MacBook Pro M2 Max
64GB
macOS
FP16
·
~61 tok/s
·
6.6 GB
Runs great
💻
MacBook Pro M3 Max
36GB
macOS
FP16
·
~61 tok/s
·
6.6 GB
Runs great
💻
MacBook Pro M3 Max
96GB
macOS
FP16
·
~61 tok/s
·
6.6 GB
Runs great
🖥️
Gaming PC (RTX 4080)
32GB
Windows
FP16
·
~109 tok/s
·
6.6 GB
Runs great
🖥️
Gaming PC (RTX 4090)
64GB
Windows
FP16
·
~153 tok/s
·
6.6 GB
Runs great
🖥️
Mac Studio M4 Max
64GB
macOS
FP16
·
~83 tok/s
·
6.6 GB
Runs great
🖥️
Mac Pro M2 Ultra
192GB
macOS
FP16
·
~121 tok/s
·
6.6 GB
Runs great
💻
MacBook Pro M4 Pro
36GB
macOS
FP16
·
~41 tok/s
·
6.6 GB
Runs great
🤖
Atom 1
64GB
Linux
FP16
·
~41 tok/s
·
6.6 GB
Runs great
🤖
Atom 1
128GB
Linux
FP16
·
~41 tok/s
·
6.6 GB
Runs great
🖥️
Mac Mini M4 Pro
24GB
macOS
FP16
·
~41 tok/s
·
6.6 GB
Runs great
🖥️
Mac Mini M4 Pro
48GB
macOS
FP16
·
~41 tok/s
·
6.6 GB
Runs great
🖥️
Gaming PC (RTX 4070)
32GB
Windows
FP16
·
~76 tok/s
·
6.6 GB
Runs well
🖥️
Gaming PC (RTX 3060)
32GB
Windows
FP16
·
~55 tok/s
·
6.6 GB
Runs well
💻
MacBook Pro M1 Pro
16GB
macOS
FP16
·
~30 tok/s
·
6.6 GB
Runs great
💻
MacBook Pro M1 Pro
32GB
macOS
FP16
·
~30 tok/s
·
6.6 GB
Runs great
💻
MacBook Pro M2 Pro
16GB
macOS
FP16
·
~30 tok/s
·
6.6 GB
Runs great
💻
MacBook Pro M2 Pro
32GB
macOS
FP16
·
~30 tok/s
·
6.6 GB
Runs great
🤖
Atom 1
32GB
Linux
FP16
·
~31 tok/s
·
6.6 GB
Runs great
🖥️
Mac Mini M2 Pro
16GB
macOS
FP16
·
~30 tok/s
·
6.6 GB
Runs great
🖥️
Mac Mini M2 Pro
32GB
macOS
FP16
·
~30 tok/s
·
6.6 GB
Runs great
💻
MacBook Pro M3 Pro
18GB
macOS
FP16
·
~23 tok/s
·
6.6 GB
Runs great
💻
MacBook Pro M3 Pro
36GB
macOS
FP16
·
~23 tok/s
·
6.6 GB
Runs great
💻
MacBook Air M4
16GB
macOS
FP16
·
~18 tok/s
·
6.6 GB
Runs great
🖥️
Mac Mini M4
16GB
macOS
FP16
·
~18 tok/s
·
6.6 GB
Runs great
🖥️
Mac Mini M4
32GB
macOS
FP16
·
~18 tok/s
·
6.6 GB
Runs great
💻
MacBook Air M3
16GB
macOS
FP16
·
~15 tok/s
·
6.6 GB
Runs great
💻
MacBook Air M2
8GB
macOS
Q8
·
~28 tok/s
·
3.6 GB
Runs well
🖥️
Mac Mini M2
8GB
macOS
Q8
·
~28 tok/s
·
3.6 GB
Runs well
🎮
Steam Deck OLED
16GB
Linux
FP16
·
~13 tok/s
·
6.6 GB
Runs great
💻
MacBook Air M1
16GB
macOS
FP16
·
~10 tok/s
·
6.6 GB
Runs great
💻
MacBook Pro M1
16GB
macOS
FP16
·
~10 tok/s
·
6.6 GB
Runs great
🖥️
Mac Mini M1
16GB
macOS
FP16
·
~10 tok/s
·
6.6 GB
Runs great
💻
Snapdragon X Elite Laptop
16GB
Windows
FP16
·
~21 tok/s
·
6.6 GB
Runs well
🍓
Raspberry Pi 5
8GB
Linux
Q8
·
~9 tok/s
·
3.6 GB
Runs great
💻
MacBook Air M1
8GB
macOS
Q8
·
~19 tok/s
·
3.6 GB
Runs well
🖥️
Mac Mini M1
8GB
macOS
Q8
·
~19 tok/s
·
3.6 GB
Runs well
📱
Galaxy S25 Ultra
12GB
Android
Q8
·
~15 tok/s
·
3.6 GB
Runs well
📱
Galaxy S24
8GB
Android
Q6
·
~16 tok/s
·
2.6 GB
Runs well
📱
iPad Pro M4
16GB
iOS
FP16
·
~13 tok/s
·
6.6 GB
Tight fit
📱
iPhone 16 Pro
8GB
iOS
Q8
·
~13 tok/s
·
3.6 GB
Tight fit
📱
OnePlus 13
16GB
Android
FP16
·
~8 tok/s
·
6.6 GB
Tight fit
📱
iPhone 15
6GB
iOS
Q6
·
~11 tok/s
·
2.6 GB
Tight fit
📱
Pixel 9 Pro
16GB
Android
FP16
·
~7 tok/s
·
6.6 GB
Tight fit
COMPARE WITH ANOTHER MODEL
Llama 3.2 3B vs Qwen 2.5 3B
Side-by-side compatibility comparison
→
RUN WITH THESE APPS
Beginner
Free
LM Studio
Beautiful desktop app for running local AI models
Tinkerer
Free
Ollama
Lightweight CLI tool for running models in the background
Beginner
Free
Jan
Open-source ChatGPT-like desktop app
Beginner
Free
LocallyAI
Run AI models privately on your iPhone and iPad
HOW TO RUN
Step-by-step run guides coming soon — check the apps above to get started today.