Back to leaderboard

qwen3.8-27b-mlx

LM-STUDIO MLX

qwen3_5 · 27B · 8bit

Marginal

Aug 17, 2026 · Apple M5 Max

qwen3.6-35b-a3b

LM-STUDIO GGUF

qwen35moe · 35B · Q4_K_S

Not Rec.

Apr 20, 2026 · Apple M2 Max

Global Score
59 vs 82
Hardware Fit
63 vs 80
Quality Score
58 vs 83

Hardware

qwen3.8-27b-mlx qwen3.6-35b-a3b
MachineMacBook ProMac Studio
CPUApple M5 MaxApple M2 Max
Cores1812
RAM64 GB96 GB
GPUApple M5 MaxApple M2 Max
OSmacOS 27.0macOS 26.1
Archarm64arm64
Power Modebalancedbalanced

Performance

qwen3.8-27b-mlx qwen3.6-35b-a3b
Tokens/sec18.354.7
First chunk4 ms32 ms
TTFT530 ms30.0 s
Load timeN/AN/A
Memory usage38.5 GB30.3 GB
Memory %60%32%

HW Fit Score Breakdown

qwen3.8-27b-mlx

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

qwen3.6-35b-a3b

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

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

qwen3.6-35b-a3b

Reasoning
18/20
Coding
10/20
Instruction
18/20
Structured
14/15
Math
14/15
Multilingual
9/10
Reasoning: Strong Coding: Weak 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