qwen3.8-27b-mlx
LM-STUDIO MLXqwen3_5 · 27B · 4bit
Oct 20, 2026 · Apple M1 Pro
qwen3.5:4b
OLLAMA GGUFqwen35 · 4.7B · Q4_K_M
Mar 6, 2026 · AMD RYZEN AI MAX+ 395 w/ Radeon 8060S
Global Score
78 vs 82
Hardware Fit
56 vs 93
Quality Score
87 vs 77
Hardware
qwen3.8-27b-mlx qwen3.5:4b
MachineMacBook ProAZW GTR Pro
CPUApple M1 ProAMD RYZEN AI MAX+ 395 w/ Radeon 8060S
Cores1032
RAM32 GB125 GB
GPUApple M1 ProIntegrated / Apple Silicon
OSmacOS 15.6.1Ubuntu 24.04.4 LTS
Archarm64x64
Power Modebalancedperformance
Performance
qwen3.8-27b-mlx qwen3.5:4b
Tokens/sec10.846.8
First chunk8 ms174 ms
TTFT2.6 s174 ms
Load timeN/A5.9 s
Memory usage21.0 GB16.2 GB
Memory %66%13%
HW Fit Score Breakdown
qwen3.8-27b-mlx
Speed
23/50
TTFT
16/20
Memory
17/30
qwen3.5:4b
Speed
43/50
TTFT
20/20
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
Reasoning
15/20
Coding
14/20
Instruction
16/20
Structured
15/15
Math
8/15
Multilingual
9/10
Reasoning: Strong Coding: Adequate Instruction Following: Strong Structured Output: Strong 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