qwen2.5:0.5b

qwen2 · 494.03M · Q4_K_M

Good

Mar 2, 2026 · Apple M4

qwen3:8b

qwen3 · 8.2B · Q4_K_M

Marginal

Mar 2, 2026 · Apple M4

Global Score

63 vs 50

Hardware Fit

100 vs 67

Quality Score

39 vs 38

Hardware

qwen2.5:0.5b qwen3:8b

MachineMacBook AirMacBook Air

CPUApple M4Apple M4

Cores1010

RAM32 GB32 GB

GPUApple M4Apple M4

OSmacOS 26.3macOS 26.3

Archarm64arm64

Power Modebalancedbalanced

Performance

qwen2.5:0.5b qwen3:8b

Tokens/sec165.018.7

TTFT81 ms12.1 s

Load time0.5 s1.2 s

Memory usage1.7 GB9.5 GB

Memory %5%30%

HW Fit Score Breakdown

qwen2.5:0.5b

Speed

40/40

TTFT

30/30

Memory

30/30

qwen3:8b

Speed

31/40

TTFT

6/30

Memory

30/30

Quality

qwen2.5:0.5b

Reasoning

6/20

Coding

6/20

Instruction

7/20

Structured

11/15

Math

5/15

Multilingual

4/10

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

qwen3:8b

Reasoning

5/20

Coding

5/20

Instruction

4/20

Structured

12/15

Math

2/15

Multilingual

10/10

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

View qwen2.5:0.5b details | View qwen3:8b details

Run yours and compare

npx metrillm@latest bench

Requires Node 20+ and Ollama or LM Studio running