qwen3:14b
OLLAMA GGUFqwen3 · 14.8B · Q4_K_M
Jun 13, 2026 · AMD Ryzen 9 3900X 12-Core Processor
qwen3:30b-a3b
OLLAMA GGUFqwen3moe · 30.5B · Q4_K_M
Jun 25, 2026 · Cortex-X925
Global Score
82 vs 84
Hardware Fit
90 vs 93
Quality Score
79 vs 80
Hardware
qwen3:14b qwen3:30b-a3b
MachineMicro-Star International Co., Ltd. MS-7C37NVIDIA NVIDIA_DGX_Spark
CPUAMD Ryzen 9 3900X 12-Core ProcessorCortex-X925
Cores2420
RAM64 GB122 GB
GPUGameViewer Virtual Display Adapter, NVIDIA GeForce RTX 3080Device 2e12
OSMicrosoft Windows 11 专业版 10.0.26200Ubuntu 24.04.4 LTS
Archx64arm64
Power Modebalancedperformance
Performance
qwen3:14b qwen3:30b-a3b
Tokens/sec74.492.7
First chunk324 ms286 ms
TTFT2.6 s1.5 s
Load time5.1 s15.3 s
Memory usage9.0 GB20.3 GB
Memory %14%17%
HW Fit Score Breakdown
qwen3:14b
Speed
50/50
TTFT
10/20
Memory
30/30
qwen3:30b-a3b
Speed
50/50
TTFT
13/20
Memory
30/30
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
qwen3:30b-a3b
Reasoning
16/20
Coding
10/20
Instruction
18/20
Structured
13/15
Math
13/15
Multilingual
10/10
Reasoning: Strong Coding: Adequate 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