Back to leaderboard

qwen3.8-27b-mlx

LM-STUDIO MLX

qwen3_5 · 27B · 8bit

Marginal

Aug 17, 2026 · Apple M5 Max

qwen/qwen3-30b-a3b-2507

LM-STUDIO GGUF

qwen3moe · 30B · Q3_K_L

Excellent

Jul 5, 2026 · AMD Ryzen 7 7800X3D 8-Core Processor

Global Score
59 vs 89
Hardware Fit
63 vs 90
Quality Score
58 vs 89

Hardware

qwen3.8-27b-mlx qwen/qwen3-30b-a3b…
MachineMacBook ProASUS
CPUApple M5 MaxAMD Ryzen 7 7800X3D 8-Core Processor
Cores1816
RAM64 GB31 GB
GPUApple M5 MaxNVIDIA GeForce RTX 4090, Raphael
OSmacOS 27.0cachyos rolling
Archarm64x64
Power Modebalancedperformance

Performance

qwen3.8-27b-mlx qwen/qwen3-30b-a3b…
Tokens/sec18.3151.8
First chunk4 ms8 ms
TTFT530 ms74 ms
Load timeN/AN/A
Memory usage38.5 GB18.5 GB
Memory %60%61%

HW Fit Score Breakdown

qwen3.8-27b-mlx

Speed
23/50
TTFT
20/20
Memory
20/30

qwen/qwen3-30b-a3b-2507

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

Quality

qwen3.8-27b-mlx

Reasoning
19/20
Coding
14/20
Instruction
6/20
Structured
1/15
Math
13/15
Multilingual
5/10
Reasoning: Strong Coding: Adequate Instruction Following: Weak Structured Output: Poor Math: Strong Multilingual: Weak

qwen/qwen3-30b-a3b-2507

Reasoning
18/20
Coding
19/20
Instruction
16/20
Structured
15/15
Math
11/15
Multilingual
10/10
Reasoning: Strong Coding: Strong Instruction Following: Strong Structured Output: Strong Math: Adequate Multilingual: Strong
View qwen3.8-27b-mlx details | View qwen/qwen3-30b-a3b… 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