Back to leaderboard

qwen3.5-4b-mlx

LM-STUDIO MLX

qwen3_5 · 4B · 4bit

Not Rec.

Mar 6, 2026 · Apple M4 Pro

qwen/qwen3.6-35b-a3b

LM-STUDIO MLX

qwen3_5_moe · 35B · 4bit

Excellent

Jul 17, 2026 · Apple M5 Max

Global Score
62 vs 89
Hardware Fit
35 vs 97
Quality Score
73 vs 86

Hardware

qwen3.5-4b-mlx qwen/qwen3.6-35b-a3b
MachineMac miniMacBook Pro
CPUApple M4 ProApple M5 Max
Cores1418
RAM64 GB64 GB
GPUApple M4 ProApple M5 Max
OSmacOS 15.7.4macOS 27.0
Archarm64arm64
Power Modebalancedbalanced

Performance

qwen3.5-4b-mlx qwen/qwen3.6-35b-a3b
Tokens/sec0.0130.2
First chunk5548 ms5 ms
TTFT5.5 s135 ms
Load timeN/AN/A
Memory usage2.9 GB26.6 GB
Memory %5%42%

HW Fit Score Breakdown

qwen3.5-4b-mlx

Speed
0/50
TTFT
6/20
Memory
29/30

qwen/qwen3.6-35b-a3b

Speed
50/50
TTFT
20/20
Memory
27/30

Quality

qwen3.5-4b-mlx

Reasoning
17/20
Coding
17/20
Instruction
9/20
Structured
8/15
Math
13/15
Multilingual
9/10
Reasoning: Strong Coding: Strong Instruction Following: Weak Structured Output: Adequate Math: Strong Multilingual: Strong

qwen/qwen3.6-35b-a3b

Reasoning
15/20
Coding
18/20
Instruction
16/20
Structured
15/15
Math
12/15
Multilingual
10/10
Reasoning: Adequate Coding: Strong Instruction Following: Strong Structured Output: Strong Math: Strong Multilingual: Strong

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