Back to leaderboard

qwen3.8-27b-mlx

LM-STUDIO MLX

qwen3_5 · 27B · 8bit

Marginal

Aug 17, 2026 · Apple M5 Max

liquid/lfm2-24b-a2b

LM-STUDIO MLX

lfm2_moe · 24B · 4bit

Good

Mar 19, 2026 · Apple M4

Global Score
59 vs 74
Hardware Fit
63 vs 84
Quality Score
58 vs 69

Hardware

qwen3.8-27b-mlx liquid/lfm2-24b-a2b
MachineMacBook ProMac mini
CPUApple M5 MaxApple M4
Cores1810
RAM64 GB24 GB
GPUApple M5 MaxApple M4
OSmacOS 27.0macOS 26.3
Archarm64arm64
Power Modebalancedbalanced

Performance

qwen3.8-27b-mlx liquid/lfm2-24b-a2b
Tokens/sec18.358.9
First chunk4 ms28 ms
TTFT530 ms481 ms
Load timeN/AN/A
Memory usage38.5 GB17.5 GB
Memory %60%73%

HW Fit Score Breakdown

qwen3.8-27b-mlx

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

liquid/lfm2-24b-a2b

Speed
50/50
TTFT
20/20
Memory
14/30

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

liquid/lfm2-24b-a2b

Reasoning
11/20
Coding
12/20
Instruction
16/20
Structured
13/15
Math
8/15
Multilingual
9/10
Reasoning: Adequate Coding: Adequate Instruction Following: Strong 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