Anthropic releases Claude Opus 5.5: 40% cheaper than Opus 5, matches Fable 5.1

addyosmani · x · 2026-09-23

Anthropic launched Claude Opus 5.5, first of the Claude 5.5 family, matching Claude Fable 5.1 on most tasks while costing 40% less to run than Opus 5, with cache reads 60% cheaper.
Performance: one tester completed a 680,000-line code migration in under a day; it cut web app load times in 39 of 40 attempts (Opus 5 made smaller, behavior-breaking improvements); it topped a single-prompt game-building comparison on graphics and polish.
Safety: best scores to date on Anthropic's automated behavioral audit — less likely to take hard-to-reverse actions, better at staying in bounds, and more resistant to prompt injection than Opus 5. Externally evaluated by Frontier Design and METR; it's the first release since Anthropic called for pacing the frontier. Addy Osmani calls it a big step up for agentic coding and computer use.

More from Models