qwen3.5-4b-mlx
LM-STUDIO MLXqwen3_5 · 4B · 4bit
Mar 3, 2026 · Apple M4
qwen2.5:3b
OLLAMA GGUFqwen2 · 3.1B · Q4_K_M
Jun 26, 2026 · Cortex-A78AE
Global Score
71 vs 77
Hardware Fit
88 vs 100
Quality Score
63 vs 67
Hardware
qwen3.5-4b-mlx qwen2.5:3b
MachineMacBook AirNVIDIA NVIDIA Jetson Orin Nano Engineering Reference Developer Kit Super
CPUApple M4Cortex-A78AE
Cores106
RAM32 GB7 GB
GPUApple M4Integrated / Unknown
OSmacOS 26.3Ubuntu 22.04.5 LTS
Archarm64arm64
Power Modebalancedbalanced
Performance
qwen3.5-4b-mlx qwen2.5:3b
Tokens/sec23.121.4
First chunk3279 ms1006 ms
TTFT3.3 s1.0 s
Load timeN/A6.4 s
Memory usage2.9 GB2.0 GB
Memory %9%27%
HW Fit Score Breakdown
qwen3.5-4b-mlx
Speed
45/50
TTFT
13/20
Memory
30/30
qwen2.5:3b
Speed
50/50
TTFT
20/20
Memory
30/30
Quality
qwen3.5-4b-mlx
Reasoning
15/20
Coding
16/20
Instruction
6/20
Structured
6/15
Math
11/15
Multilingual
9/10
Reasoning: Adequate Coding: Strong Instruction Following: Weak Structured Output: Weak Math: Adequate Multilingual: Strong
qwen2.5:3b
Reasoning
12/20
Coding
14/20
Instruction
11/20
Structured
15/15
Math
6/15
Multilingual
9/10
Reasoning: Adequate Coding: Adequate Instruction Following: Adequate Structured Output: Strong Math: Weak Multilingual: Strong
Run yours and compare
$
npm install -g metrillm@latest$
metrillmRequires Node 20+ and Ollama or LM Studio running
Or run without installing: npx metrillm@latest