qwen3.8-27b-mlx
LM-STUDIO MLXqwen3_5 · 27B · 4bit
Oct 20, 2026 · Apple M1 Pro
qwen3.5:4b-mlx
OLLAMA GGUFnvfp4
Aug 25, 2026 · Apple M1 Pro
Global Score
78 vs 68
Hardware Fit
56 vs 91
Quality Score
87 vs 58
Hardware
qwen3.8-27b-mlx qwen3.5:4b-mlx
MachineMacBook ProMacBook Pro
CPUApple M1 ProApple M1 Pro
Cores1010
RAM32 GB16 GB
GPUApple M1 ProApple M1 Pro
OSmacOS 15.6.1macOS 26.6.2
Archarm64arm64
Power Modebalancedbalanced
Performance
qwen3.8-27b-mlx qwen3.5:4b-mlx
Tokens/sec10.842.2
First chunk8 ms201 ms
TTFT2.6 s4.4 s
Load timeN/A0.0 s
Memory usage21.0 GB3.8 GB
Memory %66%24%
HW Fit Score Breakdown
qwen3.8-27b-mlx
Speed
23/50
TTFT
16/20
Memory
17/30
qwen3.5:4b-mlx
Speed
50/50
TTFT
11/20
Memory
30/30
Quality
qwen3.8-27b-mlx
Reasoning
18/20
Coding
18/20
Instruction
15/20
Structured
15/15
Math
11/15
Multilingual
10/10
Reasoning: Strong Coding: Strong Instruction Following: Strong Structured Output: Strong Math: Adequate Multilingual: Strong
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