hf.co/empero-ai/Qwen3.8-4B-Distill-GGUF:Q4_K_M
OLLAMA GGUFqwen35 · 4.33B · unknown
Aug 28, 2026 · Apple M4
qwen3.5:4b-mlx
OLLAMA GGUFnvfp4
Aug 25, 2026 · Apple M1 Pro
Global Score
66 vs 68
Hardware Fit
100 vs 91
Quality Score
52 vs 58
Hardware
hf.co/empero-ai/Qw… qwen3.5:4b-mlx
MachineMacBook ProMacBook Pro
CPUApple M4Apple M1 Pro
Cores1010
RAM16 GB16 GB
GPUApple M4Apple M1 Pro
OSmacOS 26.6.1macOS 26.6.2
Archarm64arm64
Power Modebalancedbalanced
Performance
hf.co/empero-ai/Qw… qwen3.5:4b-mlx
Tokens/sec26.542.2
First chunk447 ms201 ms
TTFT447 ms4.4 s
Load time3.3 s0.0 s
Memory usage4.8 GB3.8 GB
Memory %30%24%
HW Fit Score Breakdown
hf.co/empero-ai/Qwen3.8…
Speed
50/50
TTFT
20/20
Memory
30/30
qwen3.5:4b-mlx
Speed
50/50
TTFT
11/20
Memory
30/30
Quality
hf.co/empero-ai/Qwen3.8…
Reasoning
18/20
Coding
14/20
Instruction
4/20
Structured
1/15
Math
15/15
Multilingual
0/10
Reasoning: Strong Coding: Adequate Instruction Following: Poor Structured Output: Poor Math: Strong Multilingual: Poor
qwen3.5:4b-mlx
Reasoning
16/20
Coding
2/20
Instruction
13/20
Structured
7/15
Math
11/15
Multilingual
9/10
Reasoning: Strong Coding: Poor Instruction Following: Adequate Structured Output: Weak Math: Adequate 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