qwen3.8-27b-claude-opus-reasoning-distilled
LM-STUDIO GGUFqwen35 · 27B · Q4_K_M
Aug 20, 2026 · Apple M4 Max
qwen3.5-4b-mlx
LM-STUDIO MLXqwen3_5 · 4B · 4bit
Mar 3, 2026 · Apple M4
Global Score
80 vs 71
Hardware Fit
63 vs 88
Quality Score
87 vs 63
Hardware
qwen3.8-27b-claude… qwen3.5-4b-mlx
MachineMacBook ProMacBook Air
CPUApple M4 MaxApple M4
Cores1610
RAM128 GB32 GB
GPUApple M4 MaxApple M4
OSmacOS 26.5.2macOS 26.3
Archarm64arm64
Power Modebalancedbalanced
Performance
qwen3.8-27b-claude… qwen3.5-4b-mlx
Tokens/sec24.023.1
First chunk20 ms3279 ms
TTFT4.5 s3.3 s
Load timeN/AN/A
Memory usage15.7 GB2.9 GB
Memory %12%9%
HW Fit Score Breakdown
qwen3.8-27b-claude-opus…
Speed
27/50
TTFT
6/20
Memory
30/30
qwen3.5-4b-mlx
Speed
45/50
TTFT
13/20
Memory
30/30
Quality
qwen3.8-27b-claude-opus…
Reasoning
19/20
Coding
18/20
Instruction
11/20
Structured
15/15
Math
14/15
Multilingual
10/10
Reasoning: Strong Coding: Strong Instruction Following: Adequate Structured Output: Strong Math: Strong Multilingual: Strong
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
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