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 )
qwen2.5:14b
OLLAMA GGUFqwen2 · 14.8B · Q4_K_M
Mar 1, 2026 · Apple M4
Global Score
93 vs 83
Hardware Fit
91 vs 80
Quality Score
94 vs 84
Hardware
GPT-20B:20b qwen2.5:14b
MachineHP HyperX OMEN Transcend Gaming Laptop 14-ga0xxx5CD612DXMLMacBook Air
CPUNVIDIA RTX Spark N1X (6144-core GPU, 20-core ) RTX Spark N1X (6144-core GPU, 20-core )Apple M4
Cores2010
RAM55 GB32 GB
GPUNVIDIA RTX Spark N1X (6144-core Blackwell RTX GPU), NVIDIA NPUApple M4
OSMicrosoft Windows 11 Pro 10.0.28000macOS 26.3
Archarm64arm64
Power Modebalancedbalanced
Performance
GPT-20B:20b qwen2.5:14b
Tokens/sec57.811.2
First chunk203 msN/A
TTFT3.0 s539 ms
Load time19.2 s1.0 s
Memory usage12.0 GB16.5 GB
Memory %22%52%
HW Fit Score Breakdown
GPT-20B:20b
Speed
50/50
TTFT
11/20
Memory
30/30
qwen2.5:14b
Speed
20/50
TTFT
30/20
Memory
30/30
Quality
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
qwen2.5:14b
Reasoning
17/20
Coding
18/20
Instruction
14/20
Structured
15/15
Math
10/15
Multilingual
10/10
Reasoning: Strong Coding: Strong Instruction Following: Adequate Structured Output: Strong Math: Adequate 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