phi4-mini

Not Rec.

Mar 1, 2026 · Apple M4

deepseek-v2:16b

deepseek2 · 15.7B · Q4_0

Good

Mar 2, 2026 · Apple M4

Global Score

38 vs 73

Hardware Fit

64 vs 100

Quality Score

21 vs 55

Hardware

phi4-mini deepseek-v2:16b

Machine—MacBook Air

CPUApple M4Apple M4

Cores1010

RAM32 GB32 GB

GPUApple M4Apple M4

OSmacOS 26.3macOS 26.3

Archarm64arm64

Power Modebalancedbalanced

Performance

phi4-mini deepseek-v2:16b

Tokens/sec21.756.6

TTFT11.9 s239 ms

Load time3.5 s1.2 s

Memory usage0.0 GB17.7 GB

Memory %0%55%

HW Fit Score Breakdown

phi4-mini

Speed

31/40

TTFT

3/30

Memory

30/30

deepseek-v2:16b

Speed

40/40

TTFT

30/30

Memory

30/30

Quality

phi4-mini

Reasoning

11/20

Coding

0/20

Instruction

0/20

Structured

0/15

Math

5/15

Multilingual

5/10

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

deepseek-v2:16b

Reasoning

7/20

Coding

9/20

Instruction

13/20

Structured

10/15

Math

8/15

Multilingual

8/10

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

View phi4-mini details | View deepseek-v2:16b details

Run yours and compare

npx metrillm@latest bench

Requires Node 20+ and Ollama or LM Studio running