hard-numbers
AI ModelsLanguagesFrameworksEvaluationInfrastructureDev Tools
← Back to home
Pillar

Evaluation

Aug 29, 2026

Qwen3 14B vs Llama 3.2 3B on Apple Silicon M2: An Honest Benchmark

We ran Qwen3 14B and Llama 3.2 3B on an Apple M2 with 24 GB unified memory. The 3B model is 3.5x faster, uses 4x less memory, and on a smal…

9 min read
hard-numbers

Hard numbers for engineering decisions. We don't tell you what should work — we test what actually works.

We measure, then we publish.

Topics

  • AI Models
  • Languages
  • Frameworks
  • Evaluation

More

  • About
  • Contact
  • Privacy
  • RSS
  • GitHub

© 2026 Hard Numbers. MIT licensed.

@hardnumbers