qwen3.8-27b-mlx
LM-STUDIO MLXqwen3_5 · 27B · 4bit
Oct 20, 2026 · Apple M1 Pro
qwen3-30b-a3b-thinking-2507-claude-4.5-sonnet-high-reasoning-distill-mlx
LM-STUDIO MLXqwen3_moe · 30B · MXFP4
Apr 20, 2026 · Apple M2 Max
Global Score
78 vs 97
Hardware Fit
56 vs 95
Quality Score
87 vs 98
Hardware
qwen3.8-27b-mlx qwen3-30b-a3b-thin…
MachineMacBook ProMac Studio
CPUApple M1 ProApple M2 Max
Cores1012
RAM32 GB96 GB
GPUApple M1 ProApple M2 Max
OSmacOS 15.6.1macOS 26.1
Archarm64arm64
Power Modebalancedbalanced
Performance
qwen3.8-27b-mlx qwen3-30b-a3b-thin…
Tokens/sec10.879.6
First chunk8 ms11 ms
TTFT2.6 s1.7 s
Load timeN/A7.1 s
Memory usage21.0 GB0.0 GB
Memory %66%0%
HW Fit Score Breakdown
qwen3.8-27b-mlx
Speed
23/50
TTFT
16/20
Memory
17/30
qwen3-30b-a3b-thinking-…
Speed
50/50
TTFT
15/20
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
qwen3-30b-a3b-thinking-…
Reasoning
19/20
Coding
19/20
Instruction
20/20
Structured
15/15
Math
15/15
Multilingual
10/10
Reasoning: Strong Coding: Strong Instruction Following: Strong Structured Output: Strong Math: Strong 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