Anthropic Launches Claude Opus 5.5 at 40% Lower Cost, First Model Since CEO's Slowdown Call

Anthropic released Claude Opus 5.5 on Tuesday, September 22, 2026, its first new model since CEO Dario Amodei publicly urged the industry to slow advanced AI development. The AI startup said Opus 5.5 matches Claude Fable 5.1 on most tasks while cutting usage costs 40% versus Opus 5. Pricing drops to $4 per million input tokens and $20 per million output tokens, from $5 and $25 under Opus 5.0. Anthropic said the Claude 5.5 family has completed extensive internal alignment testing and external evaluations, with safeguards added for high-risk domains including cybersecurity and biology. The company plans to release Claude Sonnet 5.5 and Claude Haiku 5.5 in coming weeks. The launch lands amid intensifying AI existential-risk debate. Former researcher Jacob Coxon resigned this month, saying OpenAI and Anthropic take AI safety insufficiently seriously, while alignment science lead Evan Hubinger put the probability of AI "destroying humanity" within a decade above 10%.