AIMicrosoft DevBlogsabout 2 hours ago
Not all model upgrades are upgrades
7 min read
Claude Sonnet 5 uses 12x more tokens on architecture tasks than Sonnet 4.6, making the cheaper-per-token model 3.7x more expensive on code upgrades. Quality dropped from 90% to 78% on architecture work. Sonnet 5 completed code upgrades 100% of the time versus 60% for Sonnet 4.6.
Level
Hype check
Tap to vote and see what everyone thinks.
