qwen3:4b
OLLAMA GGUFqwen3 · 4.0B · Q4_K_M
Jun 26, 2026 · Cortex-A78AE
Gemma4-26B:26b
OLLAMA GGUFgemma4 · 25.8B · Q4_K_M
Aug 19, 2026 · AMD Ryzen 9 9955HX3D 16-Core Processor
Global Score
32 vs 91
Hardware Fit
98 vs 93
Quality Score
4 vs 90
Hardware
qwen3:4b Gemma4-26B:26b
MachineNVIDIA NVIDIA Jetson Orin Nano Engineering Reference Developer Kit SuperHP HyperX OMEN Gaming Laptop 16-ap1xxx
CPUCortex-A78AEAMD Ryzen 9 9955HX3D 16-Core Processor
Cores632
RAM7 GB31 GB
GPUIntegrated / UnknownAMD Radeon(TM) 610M, NVIDIA GeForce RTX 5070 Laptop GPU
OSUbuntu 22.04.5 LTSMicrosoft Windows 11 Home 10.0.26200
Archarm64x64
Power Modebalancedbalanced
Performance
qwen3:4b Gemma4-26B:26b
Tokens/sec17.940.5
First chunk950 ms727 ms
TTFT950 ms727 ms
Load time9.7 s25.3 s
Memory usage3.0 GB17.0 GB
Memory %40%55%
HW Fit Score Breakdown
qwen3:4b
Speed
50/50
TTFT
20/20
Memory
28/30
Gemma4-26B:26b
Speed
50/50
TTFT
20/20
Memory
23/30
Quality
qwen3:4b
Reasoning
4/20
Coding
0/20
Instruction
0/20
Structured
0/15
Math
0/15
Multilingual
0/10
Reasoning: Poor Coding: Poor Instruction Following: Poor Structured Output: Poor Math: Poor Multilingual: Poor
Gemma4-26B:26b
Reasoning
17/20
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
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