Why choose Sonnet 5 if Opus 4.8 looks better for the money?
is being questioned as a practical choice. A comparison graph appears to show giving better for the same cost.
is described as much cheaper, but the graph is measured in US dollars, not tokens. If delivers stronger results for the same amount of money, the reason to choose is not obvious.
may only look cheaper at low or medium settings, where it might reach similar for less money. The real question is which situations make the better everyday model choice.
Key points
- The comparison is about versus on cost and .
- The graph uses US dollars, not tokens, as the cost basis.
- appears to perform better for the same money.
- may only have a cost advantage at low or medium settings.
- Model choice should be based on real task results, not price alone.