mlx-community/quantized-gemma-2b-it

LM-STUDIO GGUF

Marginal

Mar 5, 2026 · Apple M4

mistralai/devstral-small-2-2512

LM-STUDIO MLX

mistral3 · 24B · 4bit

Good

Mar 4, 2026 · Apple M4

Global Score

45 vs 69

Hardware Fit

100 vs 50

Quality Score

21 vs 77

Hardware

mlx-community/quan… mistralai/devstral…

MachineMacBook AirMacBook Air

CPUApple M4Apple M4

Cores1010

RAM32 GB32 GB

GPUApple M4Apple M4

OSmacOS 26.3macOS 26.3

Archarm64arm64

Power Modebalancedbalanced

Performance

mlx-community/quan… mistralai/devstral…

Tokens/sec44.45.7

First chunk336 ms5007 ms

TTFT336 ms5.0 s

Load timeN/AN/A

Memory usage0.1 GB13.2 GB

Memory %0%41%

HW Fit Score Breakdown

mlx-community/quantized…

Speed

50/50

TTFT

20/20

Memory

30/30

mistralai/devstral-smal…

Speed

11/50

TTFT

9/20

Memory

30/30

Quality

mlx-community/quantized…

Reasoning

4/20

Coding

1/20

Instruction

9/20

Structured

4/15

Math

2/15

Multilingual

1/10

Reasoning: Poor Coding: Poor Instruction Following: Weak Structured Output: Weak Math: Poor Multilingual: Poor

mistralai/devstral-smal…

Reasoning

16/20

Coding

18/20

Instruction

13/20

Structured

14/15

Math

6/15

Multilingual

10/10

Reasoning: Strong Coding: Strong Instruction Following: Adequate Structured Output: Strong Math: Weak Multilingual: Strong

View mlx-community/quan… details | View mistralai/devstral… 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