Gemma4-31B:31B
OLLAMA GGUFgemma4 · 31.3B · Q4_K_M
Aug 18, 2026 · Intel Core™ Ultra X7 358H
hf.co/empero-ai/Qwen3.8-2B-Distill-GGUF:BF16
OLLAMA GGUFqwen35 · 1.94B · BF16
Aug 22, 2026 · Apple M1 Pro
Global Score
76 vs 66
Hardware Fit
45 vs 100
Quality Score
89 vs 52
Hardware
Gemma4-31B:31B hf.co/empero-ai/Qw…
MachineGMKtec NucBox EVO-T2SMacBook Pro
CPUIntel Core™ Ultra X7 358HApple M1 Pro
Cores1610
RAM64 GB16 GB
GPUIntel(R) Arc(TM) B390 GPUApple M1 Pro
OSMicrosoft Windows 11 Pro 10.0.26200macOS 26.6.2
Archx64arm64
Power Modebalancedbalanced
Performance
Gemma4-31B:31B hf.co/empero-ai/Qw…
Tokens/sec3.840.2
First chunk3536 ms325 ms
TTFT3.5 s325 ms
Load time19.6 s3.1 s
Memory usage19.9 GB4.7 GB
Memory %31%30%
HW Fit Score Breakdown
Gemma4-31B:31B
Speed
5/50
TTFT
10/20
Memory
30/30
hf.co/empero-ai/Qwen3.8…
Speed
50/50
TTFT
20/20
Memory
30/30
Quality
Gemma4-31B:31B
Reasoning
18/20
Coding
19/20
Instruction
15/20
Structured
14/15
Math
14/15
Multilingual
9/10
Reasoning: Strong Coding: Strong Instruction Following: Strong Structured Output: Strong Math: Strong Multilingual: Strong
hf.co/empero-ai/Qwen3.8…
Reasoning
15/20
Coding
9/20
Instruction
8/20
Structured
1/15
Math
15/15
Multilingual
4/10
Reasoning: Adequate Coding: Weak Instruction Following: Weak Structured Output: Poor Math: Strong Multilingual: Weak
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