Claude Opus 5.5 just dropped.

Fable 5.1-level performance on many tasks, at 40% lower cost than Opus 5. The price/performance race is getting wild.

YoeySep 221

I don't get why we're still using one model for every decision

Claude Opus 5.5 is out, and the expensive brain got cheaper overnight. Typical workloads cost 40% less than with Opus 5, while output is over 30% faster the new bill: $4 / 1M input tokens $20 / 1M output tokens

MikaSep 227

AI Newsreel — Sep 23

1 Opus 5.5 — ~40% cheaper 2 GPT-6 Sol/Luna — half API cost 3 Grok 4.7 — long-horizon 4 Google AX open-sourced 5 DeepMind Institute 6 Apple local AI leap Cost war + standards fight.

Ken AadamsSep 22

The cheapest frontier model just shipped.

Claude dropped Opus 5.5. Same family. Same jobs. 40% less than Opus 5 It works at Fable 5.1 level on most tasks. Codes. Runs the computer. Does the knowledge work. And it is more than 30% faster. The model is not the story. The bill is.

NoxSep 221