qwen3:14b
OLLAMA GGUFqwen3 · 14.8B · Q4_K_M
Jun 13, 2026 · AMD Ryzen 9 3900X 12-Core Processor
qwen3.5:4b
OLLAMA GGUFqwen35 · 4.7B · Q4_K_M
Aug 26, 2026 · Apple M1 Pro
Global Score
82 vs 68
Hardware Fit
90 vs 86
Quality Score
79 vs 60
Hardware
qwen3:14b qwen3.5:4b
MachineMicro-Star International Co., Ltd. MS-7C37MacBook Pro
CPUAMD Ryzen 9 3900X 12-Core ProcessorApple M1 Pro
Cores2410
RAM64 GB16 GB
GPUGameViewer Virtual Display Adapter, NVIDIA GeForce RTX 3080Apple M1 Pro
OSMicrosoft Windows 11 专业版 10.0.26200macOS 26.6.2
Archx64arm64
Power Modebalancedbalanced
Performance
qwen3:14b qwen3.5:4b
Tokens/sec74.436.5
First chunk324 ms207 ms
TTFT2.6 s6.6 s
Load time5.1 s3.2 s
Memory usage9.0 GB5.3 GB
Memory %14%33%
HW Fit Score Breakdown
qwen3:14b
Speed
50/50
TTFT
10/20
Memory
30/30
qwen3.5:4b
Speed
50/50
TTFT
7/20
Memory
29/30
Quality
qwen3:14b
Reasoning
18/20
Coding
11/20
Instruction
17/20
Structured
14/15
Math
10/15
Multilingual
9/10
Reasoning: Strong Coding: Adequate Instruction Following: Strong Structured Output: Strong Math: Adequate Multilingual: Strong
qwen3.5:4b
Reasoning
16/20
Coding
1/20
Instruction
12/20
Structured
9/15
Math
12/15
Multilingual
10/10
Reasoning: Strong Coding: Poor Instruction Following: Adequate Structured Output: Adequate 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