microsoft/phi-4-mini-reasoning

LM-STUDIO MLX

phi3 · 3.8B · 4bit

Good

Mar 3, 2026 · Apple M4

mlx-community/Llama-3.2-1B-Instruct-4bit

LM-STUDIO GGUF

Marginal

Mar 5, 2026 · Apple M4

Global Score

62 vs 57

Hardware Fit

99 vs 100

Quality Score

46 vs 39

Hardware

microsoft/phi-4-mi… mlx-community/Llam…

MachineMacBook AirMacBook Air

CPUApple M4Apple M4

Cores1010

RAM32 GB32 GB

GPUApple M4Apple M4

OSmacOS 26.3macOS 26.3

Archarm64arm64

Power Modebalancedbalanced

Performance

microsoft/phi-4-mi… mlx-community/Llam…

Tokens/sec40.3126.4

First chunk289 ms139 ms

TTFT289 ms139 ms

Load timeN/AN/A

Memory usage2.0 GB0.3 GB

Memory %6%1%

HW Fit Score Breakdown

microsoft/phi-4-mini-re…

Speed

50/50

TTFT

20/20

Memory

29/30

mlx-community/Llama-3.2…

Speed

50/50

TTFT

20/20

Memory

30/30

Quality

microsoft/phi-4-mini-re…

Reasoning

7/20

Coding

6/20

Instruction

7/20

Structured

7/15

Math

12/15

Multilingual

7/10

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

mlx-community/Llama-3.2…

Reasoning

5/20

Coding

3/20

Instruction

15/20

Structured

10/15

Math

1/15

Multilingual

5/10

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

View microsoft/phi-4-mi… details | View mlx-community/Llam… 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