The Model Selection Matrix
For boring text work the top models are roughly interchangeable.
Everyday workhorsemost tasks
Drafting, summarising, extracting. The leading models are close - pick one and learn it.
Reasoning modelAdvisor work
Strategy, analysis, hard problems. A different class - slower, deeper.
Short shelf life
Model capabilities change monthly - treat any comparison as a snapshot, not a verdict.
When to use Tool selection, when someone assumes they must pick 'the best' model.
Read the full post →