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 )
qwen/qwen3.8-27b
LM-STUDIO MLXqwen3_5 · 27B · 8bit
Aug 18, 2026 · Apple M4 Max
Global Score
93 vs 81
Hardware Fit
91 vs 68
Quality Score
94 vs 87
Hardware
GPT-20B:20b qwen/qwen3.8-27b
MachineHP HyperX OMEN Transcend Gaming Laptop 14-ga0xxx5CD612DXMLMacBook Pro
CPUNVIDIA RTX Spark N1X (6144-core GPU, 20-core ) RTX Spark N1X (6144-core GPU, 20-core )Apple M4 Max
Cores2016
RAM55 GB128 GB
GPUNVIDIA RTX Spark N1X (6144-core Blackwell RTX GPU), NVIDIA NPUApple M4 Max
OSMicrosoft Windows 11 Pro 10.0.28000macOS 26.5.2
Archarm64arm64
Power Modebalancedbalanced
Performance
GPT-20B:20b qwen/qwen3.8-27b
Tokens/sec57.816.4
First chunk203 ms5 ms
TTFT3.0 s742 ms
Load time19.2 sN/A
Memory usage12.0 GB38.5 GB
Memory %22%30%
HW Fit Score Breakdown
GPT-20B:20b
Speed
50/50
TTFT
11/20
Memory
30/30
qwen/qwen3.8-27b
Speed
18/50
TTFT
20/20
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
qwen/qwen3.8-27b
Reasoning
17/20
Coding
18/20
Instruction
16/20
Structured
15/15
Math
11/15
Multilingual
10/10
Reasoning: Strong Coding: Strong Instruction Following: Strong 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