hf.co/Edge-Quant/Nanbeige4.1-3B-Q8_0-GGUF:Q8_0
OLLAMA GGUFllama · 3.93B · unknown
Aug 22, 2026 · Apple M1 Pro
gemma-4-26b-a4b-it@iq3_xxs
LM-STUDIO GGUFgemma4 · 26B · IQ3_XXS
Jun 29, 2026 · Intel Gen Intel® Core™ i9-11900H
Global Score
52 vs 83
Hardware Fit
75 vs 58
Quality Score
42 vs 94
Hardware
hf.co/Edge-Quant/N… gemma-4-26b-a4b-it…
MachineMacBook ProDocker Container
CPUApple M1 ProIntel Gen Intel® Core™ i9-11900H
Cores1016
RAM16 GB15 GB
GPUApple M1 ProTigerLake-H GT1 [UHD Graphics]
OSmacOS 26.6.2Ubuntu 26.04 LTS
Archarm64x64
Power Modebalancedperformance
Performance
hf.co/Edge-Quant/N… gemma-4-26b-a4b-it…
Tokens/sec43.510.3
First chunk225 ms8 ms
TTFT30.0 s3.5 s
Load time2.7 s10.9 s
Memory usage8.2 GB6.4 GB
Memory %51%41%
HW Fit Score Breakdown
hf.co/Edge-Quant/Nanbei…
Speed
50/50
TTFT
0/20
Memory
25/30
gemma-4-26b-a4b-it@iq3_…
Speed
19/50
TTFT
12/20
Memory
27/30
Quality
hf.co/Edge-Quant/Nanbei…
Reasoning
13/20
Coding
1/20
Instruction
7/20
Structured
1/15
Math
13/15
Multilingual
7/10
Reasoning: Adequate Coding: Poor Instruction Following: Weak Structured Output: Poor Math: Strong Multilingual: Adequate
gemma-4-26b-a4b-it@iq3_…
Reasoning
18/20
Coding
19/20
Instruction
19/20
Structured
15/15
Math
14/15
Multilingual
9/10
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