qwen3-vl-4b-instruct-q4_k_m.gguf

Mac Studio (Apple M2 Max)

96 GB · macOS 26.1

Tested on April 21, 2026 · Submitted by Frank

Top 67% Compare

Global Score

62 /100

Good

Hardware Fit

100/100

Quality

46/100

Get this model

🦙

Ollama

ollama pull qwen3-vl-4b-instruct-q4_k_m.gguf

View on Ollama Library

ollama.com/library/qwen3-vl-4b-instruct-q4_k_m.gguf

Get it in LM Studio

Search and download models directly from the app

🤗

Find on HuggingFace

GGUF versions & quantizations

Hardware

Machine: Mac Studio
CPU: Apple M2 Max
Cores: 12 total (8 perf + 4 eff)
Frequency: 2.4 GHz
RAM: 96 GB LPDDR5
GPU: Apple M2 Max
OS: macOS 26.1
Arch: arm64
Power Mode: balanced

Performance

Tokens/sec Tokens generated per second — higher is better: 67.8
Standard deviation Variation of speed across benchmark runs — lower means more consistent: ±27.6
First chunk latency Delay before the first streamed chunk arrives from runtime: 62 ms
Time to first token How fast the model starts responding — lower is better: 64 ms
Load time Time to load the model into memory before inference: 0.0 s
Memory usage RAM consumed during inference — percentage of total system memory: 0.0 GB (0%)
Total tokens Total number of tokens generated across all benchmark prompts: 983

Score breakdown

Speed

50/50

Time to first token

20/20

Memory

30/30

Quality

Reasoning

16/20

Coding

0/20

Instruction following

12/20

Structured output

1/15

Math

8/15

Multilingual

9/10

Category levels

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

Metadata

Spec version: 0.2.1
Runtime: Openai compatible
Model format: GGUF
Hardware profile: HIGH-END
Result hash: 4b6bf2b5fcf2e091947c886e36a6e9b3fd230c5d6a947b0807a7d2522d14dd0f

Interpretation

Hardware fit: 100/100. Overall suitability: GOOD (Global 62/100). Category profile: Reasoning: Strong, Coding: Poor, Instruction Following: Adequate, Structured Output: Poor, Math: Adequate, Multilingual: Strong.

Warnings

Token speed is unstable (stddev 27.6 tok/s, mean 67.8 tok/s) — may indicate thermal throttling or memory pressure.

Bench Environment

Power: AC CPU load: avg 7% (peak 8%)

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