Back to leaderboard

qwen3-coder:30b-a3b-q4_K_M

OLLAMA GGUF

qwen3moe · 30.5B · Q4_K_M

Marginal

Jun 18, 2026 · Intel Core™ Ultra 9 285HX

qwen3.5-2b-mlx

LM-STUDIO MLX

qwen3_5 · 2B · 8bit

Good

Mar 3, 2026 · Apple M4

Global Score
52 vs 70
Hardware Fit
69 vs 93
Quality Score
45 vs 60

Hardware

qwen3-coder:30b-a3… qwen3.5-2b-mlx
MachineMicro-Star International Co., Ltd. Titan 18 HX AI A2XWJGMacBook Air
CPUIntel Core™ Ultra 9 285HXApple M4
Cores2410
RAM63 GB32 GB
GPUIntel(R) Graphics, NVIDIA GeForce RTX 5090 Laptop GPUApple M4
OSMicrosoft Windows 11 Home Single Language 10.0.26200macOS 26.3
Archx64arm64
Power Modebalancedbalanced

Performance

qwen3-coder:30b-a3… qwen3.5-2b-mlx
Tokens/sec29.432.9
First chunk593 ms3096 ms
TTFT593 ms3.1 s
Load time14.2 sN/A
Memory usage42.6 GB2.5 GB
Memory %67%8%

HW Fit Score Breakdown

qwen3-coder:30b-a3b-q4_…

Speed
33/50
TTFT
20/20
Memory
16/30

qwen3.5-2b-mlx

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

Quality

qwen3-coder:30b-a3b-q4_…

Reasoning
16/20
Coding
18/20
Instruction
0/20
Structured
0/15
Math
11/15
Multilingual
0/10
Reasoning: Strong Coding: Strong Instruction Following: Poor Structured Output: Poor Math: Strong Multilingual: Poor

qwen3.5-2b-mlx

Reasoning
11/20
Coding
9/20
Instruction
14/20
Structured
14/15
Math
3/15
Multilingual
9/10
Reasoning: Adequate Coding: Weak Instruction Following: Adequate Structured Output: Strong Math: Poor 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