qwen3.5:4b

qwen35 · 4.7B · Q4_K_M

Excellent

Mar 2, 2026 · Apple M4

llama2:7b

llama · 7B · Q4_0

Marginal

Mar 2, 2026 · Apple M4

Global Score

80 vs 55

Hardware Fit

88 vs 99

Quality Score

76 vs 36

Hardware

qwen3.5:4b llama2:7b

MachineMacBook AirMacBook Air

CPUApple M4Apple M4

Cores1010

RAM32 GB32 GB

GPUApple M4Apple M4

OSmacOS 26.3macOS 26.3

Archarm64arm64

Power Modebalancedbalanced

Performance

qwen3.5:4b llama2:7b

Tokens/sec16.625.1

TTFT369 ms228 ms

Load time1.5 s0.4 s

Memory usage6.5 GB5.8 GB

Memory %20%18%

HW Fit Score Breakdown

qwen3.5:4b

Speed

28/40

TTFT

30/30

Memory

30/30

llama2:7b

Speed

39/40

TTFT

30/30

Memory

30/30

Quality

qwen3.5:4b

Reasoning

16/20

Coding

14/20

Instruction

15/20

Structured

15/15

Math

7/15

Multilingual

9/10

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

llama2:7b

Reasoning

4/20

Coding

3/20

Instruction

13/20

Structured

8/15

Math

1/15

Multilingual

7/10

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

View qwen3.5:4b details | View llama2:7b details

Run yours and compare

npx metrillm@latest bench

Requires Node 20+ and Ollama or LM Studio running