Back to leaderboard

qwen3.5-4b-mlx

LM-STUDIO MLX

qwen3_5 · 4B · 4bit

Good

Mar 3, 2026 · Apple M4

qwen3-vl:8b

OLLAMA GGUF

qwen3vl · 8.8B · Q4_K_M

Good

Apr 28, 2026 · Intel Xeon® Gold 6230

Global Score
71 vs 72
Hardware Fit
88 vs 89
Quality Score
63 vs 64

Hardware

qwen3.5-4b-mlx qwen3-vl:8b
MachineMacBook AirHP HP Z8 G4 Workstation
CPUApple M4Intel Xeon® Gold 6230
Cores1080
RAM32 GB252 GB
GPUApple M4TU104GL [Quadro RTX 5000]
OSmacOS 26.3Ubuntu 24.04.4 LTS
Archarm64x64
Power Modebalancedlow-power

Performance

qwen3.5-4b-mlx qwen3-vl:8b
Tokens/sec23.168.2
First chunk3279 ms285 ms
TTFT3.3 s2.0 s
Load timeN/A42.4 s
Memory usage2.9 GB7.1 GB
Memory %9%3%

HW Fit Score Breakdown

qwen3.5-4b-mlx

Speed
45/50
TTFT
13/20
Memory
30/30

qwen3-vl:8b

Speed
50/50
TTFT
9/20
Memory
30/30

Quality

qwen3.5-4b-mlx

Reasoning
15/20
Coding
16/20
Instruction
6/20
Structured
6/15
Math
11/15
Multilingual
9/10
Reasoning: Adequate Coding: Strong Instruction Following: Weak Structured Output: Weak Math: Adequate Multilingual: Strong

qwen3-vl:8b

Reasoning
16/20
Coding
3/20
Instruction
10/20
Structured
11/15
Math
14/15
Multilingual
10/10
Reasoning: Strong Coding: Poor Instruction Following: Adequate Structured Output: Adequate Math: Strong Multilingual: Strong
View qwen3.5-4b-mlx details | View qwen3-vl:8b details

Run yours and compare

$ npm install -g metrillm@latest
$ metrillm

Requires Node 20+ and Ollama or LM Studio running

Or run without installing: npx metrillm@latest