deepseek-v2:16b

deepseek2 · 15.7B · Q4_0

Good

Mar 2, 2026 · Apple M4

qwen2.5:7b

qwen2 · 7.6B · Q4_K_M

Excellent

Mar 2, 2026 · Apple M4

Global Score

73 vs 85

Hardware Fit

100 vs 95

Quality Score

55 vs 79

Hardware

deepseek-v2:16b qwen2.5:7b

MachineMacBook AirMacBook Air

CPUApple M4Apple M4

Cores1010

RAM32 GB32 GB

GPUApple M4Apple M4

OSmacOS 26.3macOS 26.3

Archarm64arm64

Power Modebalancedbalanced

Performance

deepseek-v2:16b qwen2.5:7b

Tokens/sec56.622.2

TTFT239 ms285 ms

Load time1.2 s0.8 s

Memory usage17.7 GB7.6 GB

Memory %55%24%

HW Fit Score Breakdown

deepseek-v2:16b

Speed

40/40

TTFT

30/30

Memory

30/30

qwen2.5:7b

Speed

35/40

TTFT

30/30

Memory

30/30

Quality

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

qwen2.5:7b

Reasoning

14/20

Coding

18/20

Instruction

16/20

Structured

14/15

Math

8/15

Multilingual

9/10

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

View deepseek-v2:16b details | View qwen2.5:7b details

Run yours and compare

npx metrillm@latest bench

Requires Node 20+ and Ollama or LM Studio running