qwen3.8-27b-mlx
LM-STUDIO MLXqwen3_5 · 27B · 8bit
Aug 17, 2026 · Apple M5 Max
qwen3:0.6b
OLLAMA GGUFqwen3 · 751.63M · Q4_K_M
Oct 11, 2026 · AMD Ryzen 5 5500
Global Score
59 vs 74
Hardware Fit
63 vs 100
Quality Score
58 vs 63
Hardware
qwen3.8-27b-mlx qwen3:0.6b
MachineMacBook ProMicrosoft WSL
CPUApple M5 MaxAMD Ryzen 5 5500
Cores1812
RAM64 GB8 GB
GPUApple M5 MaxIntegrated / Unknown
OSmacOS 27.0Ubuntu 26.04.1 LTS
Archarm64x64
Power Modebalancedunknown
Performance
qwen3.8-27b-mlx qwen3:0.6b
Tokens/sec18.3249.5
First chunk4 ms80 ms
TTFT530 ms659 ms
Load timeN/A6.3 s
Memory usage38.5 GB0.9 GB
Memory %60%11%
HW Fit Score Breakdown
qwen3.8-27b-mlx
Speed
23/50
TTFT
20/20
Memory
20/30
qwen3:0.6b
Speed
50/50
TTFT
20/20
Memory
30/30
Quality
qwen3.8-27b-mlx
Reasoning
19/20
Coding
14/20
Instruction
6/20
Structured
1/15
Math
13/15
Multilingual
5/10
Reasoning: Strong Coding: Adequate Instruction Following: Weak Structured Output: Poor Math: Strong Multilingual: Weak
qwen3:0.6b
Reasoning
13/20
Coding
6/20
Instruction
12/20
Structured
13/15
Math
12/15
Multilingual
7/10
Reasoning: Adequate Coding: Weak Instruction Following: Adequate Structured Output: Strong Math: Strong Multilingual: Adequate
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