qwen3.5-4b-mlx
LM-STUDIO MLXqwen3_5 · 4B · 4bit
Mar 3, 2026 · Apple M4
qwen3:30b-a3b-instruct-2507-q4_K_M
OLLAMA GGUFqwen3moe · 30.5B · Q4_K_M
Jul 31, 2026 · Intel Xeon® E5-4620 0
Global Score
71 vs 71
Hardware Fit
88 vs 42
Quality Score
63 vs 84
Hardware
qwen3.5-4b-mlx qwen3:30b-a3b-inst…
MachineMacBook AirDell Inc. PowerEdge R820
CPUApple M4Intel Xeon® E5-4620 0
Cores1064
RAM32 GB256 GB
GPUApple M4Microsoft Remote Display Adapter, Microsoft Basic Display Adapter
OSmacOS 26.3Microsoft Windows Server 2022 Datacenter 10.0.20348
Archarm64x64
Power Modebalancedbalanced
Performance
qwen3.5-4b-mlx qwen3:30b-a3b-inst…
Tokens/sec23.14.0
First chunk3279 ms2070 ms
TTFT3.3 s2.1 s
Load timeN/A44.5 s
Memory usage2.9 GB17.7 GB
Memory %9%7%
HW Fit Score Breakdown
qwen3.5-4b-mlx
Speed
45/50
TTFT
13/20
Memory
30/30
qwen3:30b-a3b-instruct-…
Speed
3/50
TTFT
9/20
Memory
30/30
Quality
qwen3.5-4b-mlx
Reasoning
15/20
Coding
16/20
Instruction
6/20
Structured
6/15
Math
11/15
Multilingual
9/10
Reasoning: Adequate Coding: Strong Instruction Following: Weak Structured Output: Weak Math: Adequate Multilingual: Strong
qwen3:30b-a3b-instruct-…
Reasoning
18/20
Coding
18/20
Instruction
11/20
Structured
15/15
Math
12/15
Multilingual
10/10
Reasoning: Strong Coding: Strong Instruction Following: Adequate 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