qwen3.5-0.8b-mlx
LM-STUDIO MLXqwen3_5 · 0.8B · 4bit
Mar 3, 2026 · Apple M4
bloomz-560m
LM-STUDIO GGUFbloom · 560M · Q4_K_M
Mar 8, 2026 · Intel Core™ i5-5300U
Global Score
55 vs 30
Hardware Fit
87 vs 85
Quality Score
41 vs 7
Hardware
qwen3.5-0.8b-mlx bloomz-560m
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
qwen3.5-0.8b-mlx bloomz-560m
Tokens/sec68.212.2
First chunk3268 ms1167 ms
TTFT3.3 s1.2 s
Load timeN/AN/A
Memory usage0.6 GB0.5 GB
Memory %2%3%
HW Fit Score Breakdown
qwen3.5-0.8b-mlx
Speed
50/50
TTFT
13/20
Memory
24/30
bloomz-560m
Speed
38/50
TTFT
20/20
Memory
27/30
Quality
qwen3.5-0.8b-mlx
Reasoning
8/20
Coding
1/20
Instruction
12/20
Structured
12/15
Math
3/15
Multilingual
5/10
Reasoning: Weak Coding: Poor Instruction Following: Adequate Structured Output: Strong Math: Poor Multilingual: Adequate
bloomz-560m
Reasoning
4/20
Coding
0/20
Instruction
2/20
Structured
0/15
Math
0/15
Multilingual
1/10
Reasoning: Poor Coding: Poor Instruction Following: Poor Structured Output: Poor Math: Poor 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