🤖Claude Opus 5.5 Hits Fable-Level Scores at $4/$20 Pricing
TL;DR
Anthropic shipped Claude Opus 5.5 on Sept 22, matching or beating Fable 5.1 on agentic coding and knowledge work while cutting per-token prices 20%. It also runs over 30% faster than Opus 5.
Anthropic shipped Claude Opus 5.5 on Sept 22, matching or beating Fable 5.1 on agentic coding and knowledge work while cutting per-token prices 20%. It also runs over 30% faster than Opus 5. Sonnet 5.5 and Haiku 5.5 are promised within weeks.
Key Points
Pricing drops from $5/$25 to $4/$20 per million input/output tokens; cache reads fall about 60%
Anthropic says it beats Fable 5.1 on agentic coding, computer use, chart reading and multidisciplinary reasoning
Arrives two months after Opus 5 (July 24) and ships on Anthropic, AWS, Google Cloud and Azure
Pre-release testing by METR and Frontier Design; ships under the same bio and cyber safeguards as Fable
First Anthropic model since Dario Amodei publicly backed pacing the frontier
Why It Matters
If a mid-priced Opus now does Fable-class work, most teams can downshift their default model and bank the savings. Watch whether Sonnet 5.5 pushes that same curve lower.
Quick Facts
Comments
Be the first to comment
Enjoyed this article?
Get it daily. 7am. Free. Reads in 5 minutes.
Join 3,568 builders reading daily.