hf.co/empero-ai/Qwen3.8-2B-Distill-GGUF:BF16

OLLAMA GGUF

qwen35 · 1.94B · BF16

Good

Aug 22, 2026 · Apple M1 Pro

mistralai/devstral-small-2-2512

LM-STUDIO MLX

mistral3 · 24B · 4bit

Good

Mar 4, 2026 · Apple M4

Global Score

66 vs 69

Hardware Fit

100 vs 50

Quality Score

52 vs 77

Hardware

hf.co/empero-ai/Qw… mistralai/devstral…

MachineMacBook ProMacBook Air

CPUApple M1 ProApple M4

Cores1010

RAM16 GB32 GB

GPUApple M1 ProApple M4

OSmacOS 26.6.2macOS 26.3

Archarm64arm64

Power Modebalancedbalanced

Performance

hf.co/empero-ai/Qw… mistralai/devstral…

Tokens/sec40.25.7

First chunk325 ms5007 ms

TTFT325 ms5.0 s

Load time3.1 sN/A

Memory usage4.7 GB13.2 GB

Memory %30%41%

HW Fit Score Breakdown

hf.co/empero-ai/Qwen3.8…

Speed

50/50

TTFT

20/20

Memory

30/30

mistralai/devstral-smal…

Speed

11/50

TTFT

9/20

Memory

30/30

Quality

hf.co/empero-ai/Qwen3.8…

Reasoning

15/20

Coding

9/20

Instruction

8/20

Structured

1/15

Math

15/15

Multilingual

4/10

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

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 hf.co/empero-ai/Qw… 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