Gemma4-26B:26b
OLLAMA GGUFgemma4 · 25.8B · Q4_K_M
Aug 24, 2026 · Intel Core™ Ultra X7 358H
qwen3.5:9b
OLLAMA GGUFqwen35 · 9.7B · Q4_K_M
Mar 7, 2026 · AMD Ryzen 5 5500
Global Score
87 vs 71
Hardware Fit
82 vs 56
Quality Score
89 vs 78
Hardware
Gemma4-26B:26b qwen3.5:9b
MachineGMKtec NucBox EVO-T2SSystem manufacturer
CPUIntel Core™ Ultra X7 358HAMD Ryzen 5 5500
Cores1612
RAM64 GB16 GB
GPUIntel(R) Arc(TM) B390 GPUNVIDIA GeForce RTX 3060
OSMicrosoft Windows 11 Pro 10.0.26200Microsoft Windows 11 IoT Enterprise LTSC 10.0.26100
Archx64x64
Power Modebalancedbalanced
Performance
Gemma4-26B:26b qwen3.5:9b
Tokens/sec23.45.3
First chunk1073 ms1624 ms
TTFT1.1 s1.6 s
Load time16.0 s4.0 s
Memory usage16.7 GB7.9 GB
Memory %26%50%
HW Fit Score Breakdown
Gemma4-26B:26b
Speed
32/50
TTFT
20/20
Memory
30/30
qwen3.5:9b
Speed
11/50
TTFT
20/20
Memory
25/30
Quality
Gemma4-26B:26b
Reasoning
16/20
Coding
19/20
Instruction
17/20
Structured
14/15
Math
13/15
Multilingual
10/10
Reasoning: Strong Coding: Strong Instruction Following: Strong Structured Output: Strong Math: Strong Multilingual: Strong
qwen3.5:9b
Reasoning
14/20
Coding
17/20
Instruction
12/20
Structured
15/15
Math
10/15
Multilingual
10/10
Reasoning: Adequate 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