qwen3-30b-a3b-thinking-2507-claude-4.5-sonnet-high-reasoning-distill-mlx
LM-STUDIO MLXqwen3_moe · 30B · MXFP4
Apr 22, 2026 · Apple M2 Max
qwen3.5-0.8b-mlx
LM-STUDIO MLXqwen3_5 · 0.8B · 4bit
Mar 6, 2026 · Apple M4 Pro
Global Score
95 vs 53
Hardware Fit
94 vs 89
Quality Score
96 vs 38
Hardware
qwen3-30b-a3b-thin… qwen3.5-0.8b-mlx
MachineMac StudioMac mini
CPUApple M2 MaxApple M4 Pro
Cores1214
RAM96 GB64 GB
GPUApple M2 MaxApple M4 Pro
OSmacOS 26.1macOS 15.7.4
Archarm64arm64
Power Modebalancedbalanced
Performance
qwen3-30b-a3b-thin… qwen3.5-0.8b-mlx
Tokens/sec80.452.9
First chunk46 ms4013 ms
TTFT1.8 s4.0 s
Load timeN/AN/A
Memory usage21.2 GB0.6 GB
Memory %22%1%
HW Fit Score Breakdown
qwen3-30b-a3b-thinking-…
Speed
50/50
TTFT
14/20
Memory
30/30
qwen3.5-0.8b-mlx
Speed
50/50
TTFT
9/20
Memory
30/30
Quality
qwen3-30b-a3b-thinking-…
Reasoning
19/20
Coding
18/20
Instruction
19/20
Structured
15/15
Math
15/15
Multilingual
10/10
Reasoning: Strong Coding: Strong Instruction Following: Strong Structured Output: Strong Math: Strong Multilingual: Strong
qwen3.5-0.8b-mlx
Reasoning
7/20
Coding
2/20
Instruction
12/20
Structured
10/15
Math
2/15
Multilingual
5/10
Reasoning: Weak Coding: Poor Instruction Following: Adequate Structured Output: Adequate Math: Poor 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