qwen3.5-4b-mlx
LM-STUDIO MLXqwen3_5 · 4B · 4bit
Mar 3, 2026 · Apple M4
falcon-h1-0.5b-instruct
LM-STUDIO GGUFfalcon-h1 · 0.5B · Q4_K_M
Mar 8, 2026 · Intel Core™ i5-5300U
Global Score
71 vs 57
Hardware Fit
88 vs 91
Quality Score
63 vs 42
Hardware
qwen3.5-4b-mlx falcon-h1-0.5b-ins…
MachineMacBook AirLENOVO 20BUS00700
CPUApple M4Intel Core™ i5-5300U
Cores104
RAM32 GB16 GB
GPUApple M4Intel(R) HD Graphics 5500
OSmacOS 26.3Microsoft Windows 10 Professionnel 10.0.19045
Archarm64x64
Power Modebalancedbalanced
Performance
qwen3.5-4b-mlx falcon-h1-0.5b-ins…
Tokens/sec23.116.3
First chunk3279 ms2987 ms
TTFT3.3 s3.0 s
Load timeN/AN/A
Memory usage2.9 GB0.3 GB
Memory %9%2%
HW Fit Score Breakdown
qwen3.5-4b-mlx
Speed
45/50
TTFT
13/20
Memory
30/30
falcon-h1-0.5b-instruct
Speed
48/50
TTFT
16/20
Memory
27/30
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
falcon-h1-0.5b-instruct
Reasoning
8/20
Coding
7/20
Instruction
12/20
Structured
10/15
Math
4/15
Multilingual
1/10
Reasoning: Weak Coding: Weak Instruction Following: Adequate Structured Output: Adequate Math: Weak Multilingual: Poor
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