Back to leaderboard

qwen3.8-27b-mlx

LM-STUDIO MLX

qwen3_5 · 27B · 8bit

Marginal

Aug 17, 2026 · Apple M5 Max

gemma4:31b

OLLAMA GGUF

gemma4 · 31.3B · Q4_K_M

Not Rec.

Jul 7, 2026 · Cortex-X925

Global Score
59 vs 66
Hardware Fit
63 vs 40
Quality Score
58 vs 77

Hardware

qwen3.8-27b-mlx gemma4:31b
MachineMacBook ProASUSTeK COMPUTER INC. GX10
CPUApple M5 MaxCortex-X925
Cores1820
RAM64 GB122 GB
GPUApple M5 MaxDevice 2e12
OSmacOS 27.0Ubuntu 24.04.4 LTS
Archarm64arm64
Power Modebalancedperformance

Performance

qwen3.8-27b-mlx gemma4:31b
Tokens/sec18.310.7
First chunk4 ms983 ms
TTFT530 ms20.3 s
Load timeN/A8.8 s
Memory usage38.5 GB21.4 GB
Memory %60%18%

HW Fit Score Breakdown

qwen3.8-27b-mlx

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

gemma4:31b

Speed
10/50
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

gemma4:31b

Reasoning
14/20
Coding
17/20
Instruction
13/20
Structured
14/15
Math
10/15
Multilingual
9/10
Reasoning: Adequate Coding: Strong Instruction Following: Adequate Structured Output: Strong Math: Adequate 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