qwen3.5:27b

qwen35 · 27.8B · Q4_K_M

THINKING MODEL

Mac mini (Apple M4)

24 GB · macOS 26.3

Tested on April 1, 2026 · Submitted by cryptoepops

Top 100% Compare

Global Score

4 /100

Not Rec.

Hardware Fit

13/100

Quality

0/100

Get this model

🦙

Ollama

ollama pull qwen3.5:27b

View on Ollama Library

ollama.com/library/qwen3.5

Get it in LM Studio

Search and download models directly from the app

🤗

Find on HuggingFace

GGUF versions & quantizations

Hardware

Machine: Mac mini
CPU: Apple M4
Cores: 10 total (4 perf + 6 eff)
Frequency: 2.4 GHz
RAM: 24 GB LPDDR5
GPU: Apple M4
OS: macOS 26.3
Arch: arm64
Power Mode: balanced

Performance

Tokens/sec Tokens generated per second — higher is better: 3.5
Standard deviation Variation of speed across benchmark runs — lower means more consistent: ±0.1
First chunk latency Delay before the first streamed chunk arrives from runtime: 1.8 s
Time to first token How fast the model starts responding — lower is better: 52.0 s
Load time Time to load the model into memory before inference: 13.4 s
Memory usage RAM consumed during inference — percentage of total system memory: 21.3 GB (89%)
Total tokens Total number of tokens generated across all benchmark prompts: 1429
Thinking tokens (est.) Estimated internal reasoning tokens — models with "thinking" generate more tokens: ~722

Score breakdown

Speed

7/50

Time to first token

0/20

Memory

6/30

Quality

Reasoning

0/20

Coding

0/20

Instruction following

0/20

Structured output

0/15

Math

0/15

Multilingual

0/10

Category levels

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

Metadata

Spec version: 0.2.1
Runtime: Ollama 0.19.0
Model format: GGUF
Hardware profile: ENTRY
Result hash: d17a1a96c32d6a5f056d55c33e714138754ef141863abd4c60b700eb997503ec

Interpretation

Hardware fit: 13/100. Overall suitability: NOT RECOMMENDED (Global 4/100). Category profile: Reasoning: Poor, Coding: Poor, Instruction Following: Poor, Structured Output: Poor, Math: Poor, Multilingual: Poor. Warning: model produced very low accuracy on quality tasks — results may be unusable despite good hardware performance.

Warnings

Model produced very low accuracy on quality tasks — results may be unusable despite good hardware performance.
Significant swap activity during benchmark (+2.3 GB). Model may exceed available RAM — results are severely degraded.

Disqualifiers

Time to first token too high: 52036ms (maximum: 21386ms for ENTRY profile)

Bench Environment

Power: AC Swap delta: +2.3 GB CPU load: avg 39% (peak 42%)

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