qwen3:4b
OLLAMA GGUFqwen3 · 4.0B · Q4_K_M
Jun 26, 2026 · Cortex-A78AE
qwen/qwen3.5-35b-a3b
LM-STUDIO GGUFqwen35moe · 35B · Q6_K
Sep 19, 2026 · AMD Ryzen 7 5800X 8-Core Processor
Global Score
29 vs 72
Hardware Fit
86 vs 35
Quality Score
4 vs 88
Hardware
qwen3:4b qwen/qwen3.5-35b-a3b
MachineNVIDIA NVIDIA Jetson Orin Nano Engineering Reference Developer Kit SuperSystem manufacturer
CPUCortex-A78AEAMD Ryzen 7 5800X 8-Core Processor
Cores616
RAM7 GB32 GB
GPUIntegrated / UnknownAMD Radeon RX 9060 XT
OSUbuntu 22.04.5 LTSMicrosoft Windows 11 Home 10.0.26200
Archarm64x64
Power Modebalancedbalanced
Performance
qwen3:4b qwen/qwen3.5-35b-a3b
Tokens/sec17.98.3
First chunk1010 ms32 ms
TTFT6.4 s2.0 s
Load time9.5 sN/A
Memory usage3.0 GB28.9 GB
Memory %40%91%
HW Fit Score Breakdown
qwen3:4b
Speed
50/50
TTFT
8/20
Memory
28/30
qwen/qwen3.5-35b-a3b
Speed
13/50
TTFT
17/20
Memory
5/30
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
qwen/qwen3.5-35b-a3b
Reasoning
16/20
Coding
19/20
Instruction
16/20
Structured
15/15
Math
12/15
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