gemma4:latest

gemma4 · 8.0B · Q4_K_M

THINKING MODEL

LENOVO 82JQ (AMD Ryzen 7 5800H)

16 GB · Microsoft Windows 11 家庭版中文版 10.0.26100

Tested on June 5, 2026

Top 6% Compare

Global Score

89 /100

Excellent

Hardware Fit

83/100

Quality

92/100

Get this model

🦙

Ollama

ollama pull gemma4:latest

View on Ollama Library

ollama.com/library/gemma4

Get it in LM Studio

Search and download models directly from the app

🤗

Find on HuggingFace

GGUF versions & quantizations

Hardware

Machine: LENOVO 82JQ
CPU: AMD Ryzen 7 5800H
Cores: 16 total (16 perf)
Frequency: 3.2 GHz
RAM: 16 GB DDR4
GPU: NVIDIA GeForce RTX 3060 Laptop GPU
OS: Microsoft Windows 11 家庭版中文版 10.0.26100
Arch: x64
Power Mode: balanced

Performance

Tokens/sec Tokens generated per second — higher is better: 46.3
Standard deviation Variation of speed across benchmark runs — lower means more consistent: ±1.4
First chunk latency Delay before the first streamed chunk arrives from runtime: 824 ms
Time to first token How fast the model starts responding — lower is better: 3.8 s
Load time Time to load the model into memory before inference: 21.0 s
Memory usage RAM consumed during inference — percentage of total system memory: 8.9 GB (56%)
Total tokens Total number of tokens generated across all benchmark prompts: 1460
Thinking tokens (est.) Estimated internal reasoning tokens — models with "thinking" generate more tokens: ~622

Score breakdown

Speed

50/50

Time to first token

11/20

Memory

22/30

Quality

Reasoning

18/20

Coding

19/20

Instruction following

15/20

Structured output

15/15

Math

15/15

Multilingual

10/10

Category levels

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

Metadata

Spec version: 0.2.1
Runtime: Ollama 0.30.4
Model format: GGUF
Hardware profile: BALANCED
Result hash: beef1a5467bdbdb0cc07a18a612542d61b31ea791f790958cf2282a2e15413a3

Interpretation

Hardware fit: 83/100. Overall suitability: EXCELLENT (Global 89/100). Category profile: Reasoning: Strong, Coding: Strong, Instruction Following: Strong, Structured Output: Strong, Math: Strong, Multilingual: Strong.

Warnings

Significant swap activity during benchmark (+2.1 GB). Model may exceed available RAM — results are severely degraded.

Bench Environment

Power: AC Swap delta: +2.1 GB CPU load: avg 59% (peak 80%)

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