gemma4:12b

gemma4 · 11.9B · Q4_K_M

THINKING MODEL

MacBook Pro (Apple M1 Pro)

16 GB · macOS 26.6.2

Tested on August 22, 2026

Top 75% Compare

Global Score

63 /100

Not Rec.

Hardware Fit

62/100

Quality

63/100

Get this model

🦙

Ollama

ollama pull gemma4:12b

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: MacBook Pro
CPU: Apple M1 Pro
Cores: 10 total (8 perf + 2 eff)
Frequency: 2.4 GHz
RAM: 16 GB LPDDR5
GPU: Apple M1 Pro
OS: macOS 26.6.2
Arch: arm64
Power Mode: balanced

Performance

Tokens/sec Tokens generated per second — higher is better: 15.3
Standard deviation Variation of speed across benchmark runs — lower means more consistent: ±0.0
First chunk latency Delay before the first streamed chunk arrives from runtime: 1.1 s
Time to first token How fast the model starts responding — lower is better: 30.0 s
Load time Time to load the model into memory before inference: 5.4 s
Memory usage RAM consumed during inference — percentage of total system memory: 7.6 GB (47%)
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: ~750

Score breakdown

Speed

36/50

Time to first token

0/20

Memory

26/30

Quality

Reasoning

10/20

Coding

10/20

Instruction following

9/20

Structured output

14/15

Math

10/15

Multilingual

10/10

Category levels

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

Metadata

Spec version: 0.2.1
Runtime: Ollama 0.32.14
Model format: GGUF
Hardware profile: ENTRY
Result hash: ccad00d73cb60ba9c94a6618e2b231f35226e81dc3e7adc22ac70fbc453dbe8a

Interpretation

Hardware fit: 62/100. Overall suitability: NOT RECOMMENDED (Global 63/100). Category profile: Reasoning: Weak, Coding: Adequate, Instruction Following: Weak, Structured Output: Strong, Math: Adequate, Multilingual: Strong.

Disqualifiers

Time to first token too high: 30000ms (maximum: 22068ms for ENTRY profile)

Bench Environment

Power: AC CPU load: avg 17% (peak 29%)

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