qwen3.8-27b-mlx
LM-STUDIO MLXqwen3_5 · 27B · 8bit
Aug 17, 2026 · Apple M5 Max
mistral:latest
OLLAMA GGUFllama · 7.2B · Q4_K_M
Mar 2, 2026 · Apple M4
Global Score
59 vs 73
Hardware Fit
63 vs 96
Quality Score
58 vs 63
Hardware
qwen3.8-27b-mlx mistral:latest
MachineMacBook ProMacBook Air
CPUApple M5 MaxApple M4
Cores1810
RAM64 GB32 GB
GPUApple M5 MaxApple M4
OSmacOS 27.0macOS 26.3
Archarm64arm64
Power Modebalancedbalanced
Performance
qwen3.8-27b-mlx mistral:latest
Tokens/sec18.322.5
First chunk4 msN/A
TTFT530 ms257 ms
Load timeN/A0.6 s
Memory usage38.5 GB10.1 GB
Memory %60%32%
HW Fit Score Breakdown
qwen3.8-27b-mlx
Speed
23/50
TTFT
20/20
Memory
20/30
mistral:latest
Speed
36/50
TTFT
30/20
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
mistral:latest
Reasoning
8/20
Coding
12/20
Instruction
15/20
Structured
14/15
Math
5/15
Multilingual
9/10
Reasoning: Weak Coding: Adequate 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