qwen3.5:9b
OLLAMA GGUFqwen35 · 9.7B · Q4_K_M
Aug 22, 2026 · Apple M1 Pro
GPT-20B:20b
OLLAMA GGUFgptoss · 20.9B · MXFP4
Aug 22, 2026 · NVIDIA RTX Spark N1X (6144-core GPU, 20-core ) RTX Spark N1X (6144-core GPU, 20-core )
Global Score
58 vs 93
Hardware Fit
79 vs 91
Quality Score
49 vs 94
Hardware
qwen3.5:9b GPT-20B:20b
MachineMacBook ProHP HyperX OMEN Transcend Gaming Laptop 14-ga0xxx5CD612DXML
CPUApple M1 ProNVIDIA RTX Spark N1X (6144-core GPU, 20-core ) RTX Spark N1X (6144-core GPU, 20-core )
Cores1020
RAM16 GB55 GB
GPUApple M1 ProNVIDIA RTX Spark N1X (6144-core Blackwell RTX GPU), NVIDIA NPU
OSmacOS 26.6.2Microsoft Windows 11 Pro 10.0.28000
Archarm64arm64
Power Modebalancedbalanced
Performance
qwen3.5:9b GPT-20B:20b
Tokens/sec22.157.8
First chunk601 ms203 ms
TTFT9.1 s3.0 s
Load time4.2 s19.2 s
Memory usage7.2 GB12.0 GB
Memory %45%22%
HW Fit Score Breakdown
qwen3.5:9b
Speed
48/50
TTFT
5/20
Memory
26/30
GPT-20B:20b
Speed
50/50
TTFT
11/20
Memory
30/30
Quality
qwen3.5:9b
Reasoning
15/20
Coding
4/20
Instruction
5/20
Structured
6/15
Math
11/15
Multilingual
8/10
Reasoning: Adequate Coding: Poor Instruction Following: Weak Structured Output: Weak Math: Adequate Multilingual: Strong
GPT-20B:20b
Reasoning
19/20
Coding
17/20
Instruction
18/20
Structured
15/15
Math
15/15
Multilingual
10/10
Reasoning: Strong Coding: Strong Instruction Following: Strong 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