gemma4:31b
OLLAMA GGUFgemma4 · 30.7B · Q4_K_M
Oct 2, 2026 · AMD Ryzen 9 5950X 16-Core Processor
qwen/qwen3.5-35b-a3b
LM-STUDIO GGUFqwen35moe · 35B · Q6_K
Sep 19, 2026 · AMD Ryzen 7 5800X 8-Core Processor
Global Score
64 vs 72
Hardware Fit
41 vs 35
Quality Score
74 vs 88
Hardware
gemma4:31b qwen/qwen3.5-35b-a3b
MachineSystem manufacturerSystem manufacturer
CPUAMD Ryzen 9 5950X 16-Core ProcessorAMD Ryzen 7 5800X 8-Core Processor
Cores3216
RAM32 GB32 GB
GPUNVIDIA GeForce RTX 3090AMD Radeon RX 9060 XT
OSMicrosoft Windows 11 Pro 10.0.26300Microsoft Windows 11 Home 10.0.26200
Archx64x64
Power Modebalancedbalanced
Performance
gemma4:31b qwen/qwen3.5-35b-a3b
Tokens/sec9.58.3
First chunk1751 ms32 ms
TTFT30.0 s2.0 s
Load time33.9 sN/A
Memory usage1.4 GB28.9 GB
Memory %4%91%
HW Fit Score Breakdown
gemma4:31b
Speed
11/50
TTFT
0/20
Memory
30/30
qwen/qwen3.5-35b-a3b
Speed
13/50
TTFT
17/20
Memory
5/30
Quality
gemma4:31b
Reasoning
11/20
Coding
17/20
Instruction
10/20
Structured
15/15
Math
12/15
Multilingual
9/10
Reasoning: Adequate Coding: Strong Instruction Following: Adequate Structured Output: Strong Math: Strong Multilingual: Strong
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