qwen3.5:9b
OLLAMA GGUFqwen35 · 9.7B · Q4_K_M
Mar 7, 2026 · AMD Ryzen 5 5500
aya:8b
OLLAMA GGUFcommand-r · 8.0B · F16
Mar 1, 2026 · Apple M4
Global Score
71 vs 66
Hardware Fit
56 vs 92
Quality Score
78 vs 55
Hardware
qwen3.5:9b aya:8b
MachineSystem manufacturerMacBook Air
CPUAMD Ryzen 5 5500Apple M4
Cores1210
RAM16 GB32 GB
GPUNVIDIA GeForce RTX 3060Apple M4
OSMicrosoft Windows 11 IoT Enterprise LTSC 10.0.26100macOS 26.3
Archx64arm64
Power Modebalancedbalanced
Performance
qwen3.5:9b aya:8b
Tokens/sec5.320.0
First chunk1624 msN/A
TTFT1.6 s300 ms
Load time4.0 s0.9 s
Memory usage7.9 GB6.0 GB
Memory %50%19%
HW Fit Score Breakdown
qwen3.5:9b
Speed
11/50
TTFT
20/20
Memory
25/30
aya:8b
Speed
32/50
TTFT
30/20
Memory
30/30
Quality
qwen3.5:9b
Reasoning
14/20
Coding
17/20
Instruction
12/20
Structured
15/15
Math
10/15
Multilingual
10/10
Reasoning: Adequate Coding: Strong Instruction Following: Adequate Structured Output: Strong Math: Adequate Multilingual: Strong
aya:8b
Reasoning
10/20
Coding
6/20
Instruction
13/20
Structured
15/15
Math
2/15
Multilingual
9/10
Reasoning: Adequate Coding: Weak Instruction Following: Adequate Structured Output: Strong Math: Poor 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