qwen3.8-27b-mlx
LM-STUDIO MLXqwen3_5 · 27B · 8bit
Aug 17, 2026 · Apple M5 Max
gemma4:e4b
OLLAMA GGUFgemma4 · 8.0B · Q4_K_M
Apr 3, 2026 · AMD RYZEN AI MAX+ 395
Global Score
59 vs 84
Hardware Fit
63 vs 98
Quality Score
58 vs 78
Hardware
qwen3.8-27b-mlx gemma4:e4b
MachineMacBook ProAZW GTR Pro
CPUApple M5 MaxAMD RYZEN AI MAX+ 395
Cores1832
RAM64 GB125 GB
GPUApple M5 MaxRadeon 8060S
OSmacOS 27.0Ubuntu 24.04.4 LTS
Archarm64x64
Power Modebalancedperformance
Performance
qwen3.8-27b-mlx gemma4:e4b
Tokens/sec18.353.0
First chunk4 ms214 ms
TTFT530 ms214 ms
Load timeN/A3.1 s
Memory usage38.5 GB15.0 GB
Memory %60%12%
HW Fit Score Breakdown
qwen3.8-27b-mlx
Speed
23/50
TTFT
20/20
Memory
20/30
gemma4:e4b
Speed
48/50
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
gemma4:e4b
Reasoning
14/20
Coding
17/20
Instruction
15/20
Structured
15/15
Math
7/15
Multilingual
10/10
Reasoning: Adequate Coding: Strong Instruction Following: Strong Structured Output: Strong Math: Weak 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