qwen3.8:27b
OLLAMA GGUFqwen35 · 27.3B · Q4_K_M
Aug 15, 2026 · Cortex-X925
qwen3.5:4b
OLLAMA GGUFqwen35 · 4.7B · Q4_K_M
Mar 5, 2026 · AMD RYZEN AI MAX+ 395
Global Score
85 vs 65
Hardware Fit
79 vs 73
Quality Score
87 vs 62
Hardware
qwen3.8:27b qwen3.5:4b
MachineASUSTeK COMPUTER INC. GX10AZW GTR Pro
CPUCortex-X925AMD RYZEN AI MAX+ 395
Cores2032
RAM122 GB125 GB
GPUDevice 2e12AMD Radeon 8060S
OSUbuntu 24.04.4 LTSUbuntu 24.04.4 LTS
Archarm64x64
Power Modeperformancelow-power
Performance
qwen3.8:27b qwen3.5:4b
Tokens/sec27.445.7
First chunk593 ms180 ms
TTFT593 ms30.0 s
Load time0.3 s5.3 s
Memory usage18.2 GB16.2 GB
Memory %15%13%
HW Fit Score Breakdown
qwen3.8:27b
Speed
29/50
TTFT
20/20
Memory
30/30
qwen3.5:4b
Speed
43/50
TTFT
0/20
Memory
30/30
Quality
qwen3.8:27b
Reasoning
18/20
Coding
17/20
Instruction
16/20
Structured
15/15
Math
11/15
Multilingual
10/10
Reasoning: Strong Coding: Strong Instruction Following: Strong Structured Output: Strong Math: Adequate Multilingual: Strong
qwen3.5:4b
Reasoning
17/20
Coding
1/20
Instruction
13/20
Structured
8/15
Math
13/15
Multilingual
10/10
Reasoning: Strong Coding: Poor Instruction Following: Adequate Structured Output: Adequate 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