@bcherny the tide keeps rising. I barely reach for the lower tiers anymore. Opus is starting to feel like the new Sonnet and Fable the new Opus... what’s even left for the next Opus?
@alexalbert__ what does “it just works” actually look like? does it choose the right path from the start, or go down the wrong one and recover without you stepping in?
Across our benchmarks, the model sets a new standard.
It scores 52.6% on Terminal-Bench-Science 0.1, more than double Fable 5. On Terminal-Bench 4.0, it scores 55.8% against 42.0% for Fable 5.