Claude Opus 5.5: Fable 5.1-level performance at a 40% discount!

Anthropic released their newest model, Claude Opus 5.5. This model scores higher on Terminal-Bench 4.0 (66.4%) than Fable 5.1 (55.8%), and is also less expensive ($4/million input tokens, $20/25 output tokens vs. $5/million input tokens).

I personally use Opus 5 every day, but only use Fable 5.1 when I want to do something very complex because it’s more expensive. If Opus 5.5 lives up to its promised capabilities (which it seems to), then I might remove the “expensive”/”cheap” distinction from my workflow.