google/gemma-4-e4b

gemma4 · 7.9B · Q4_K_M

ASUSTeK COMPUTER INC. ROG Strix G16 G614PH_G614PH (AMD Ryzen 9 8940HX)

15 GB · Microsoft Windows 11 Home Single Language 10.0.26200

Tested on September 20, 2026

Top 30% Compare

Global Score

83 /100

Excellent

Hardware Fit

96/100

Quality

78/100

Get this model

Get it in LM Studio

Search and download models directly from the app

🤗

Find on HuggingFace

MLX versions & conversions

Hardware

Machine: ASUSTeK COMPUTER INC. ROG Strix G16 G614PH_G614PH
CPU: AMD Ryzen 9 8940HX
Cores: 32 total (32 perf)
Frequency: 2.4 GHz
RAM: 15 GB DDR5
GPU: NVIDIA GeForce RTX 5050 Laptop GPU, AMD Radeon(TM) 610M
OS: Microsoft Windows 11 Home Single Language 10.0.26200
Arch: x64
Power Mode: balanced

Performance

Tokens/sec Tokens generated per second — higher is better: 35.9
Standard deviation Variation of speed across benchmark runs — lower means more consistent: ±0.4
First chunk latency Delay before the first streamed chunk arrives from runtime: 9 ms
Time to first token How fast the model starts responding — lower is better: 182 ms
Load time Time to load the model into memory before inference: N/A
Memory usage RAM consumed during inference — percentage of total system memory: 6.5 GB (43%)
Total tokens Total number of tokens generated across all benchmark prompts: 1367

Score breakdown

Speed

49/50

Time to first token

20/20

Memory

27/30

Quality

Reasoning

14/20

Coding

18/20

Instruction following

15/20

Structured output

15/15

Math

6/15

Multilingual

10/10

Category levels

Reasoning: Adequate Coding: Strong Instruction Following: Strong Structured Output: Strong Math: Weak Multilingual: Strong

Metadata

Spec version: 0.2.1
Runtime: LM Studio
Model format: GGUF
Hardware profile: BALANCED
Result hash: cceda846fb0785655e4bd1685cf84a1b8ae691ff130f8c193b61a31574a6cb59

Interpretation

Hardware fit: 96/100. Overall suitability: EXCELLENT (Global 83/100). Category profile: Reasoning: Adequate, Coding: Strong, Instruction Following: Strong, Structured Output: Strong, Math: Weak, Multilingual: Strong.

Warnings

Model memory footprint is estimated via LM Studio CLI rather than measured from a fresh load.
Thermal throttling detected after benchmark (heavy). Results may be degraded — consider cooling or shorter runs.

Bench Environment

Thermal: heavy Power: AC CPU load: avg 35% (peak 43%)

Run yours now

$ npm install -g metrillm@latest

$ metrillm

Requires Node 20+ and Ollama or LM Studio running

Or run without installing: npx metrillm@latest