The new $4 per 1 million input token price for Claude Opus 5.5 reflects a 20% discount from the $5 charged for its predecessor. This flagship model performs at the level of Claude Fable 5.1 on most tasks, reducing operational costs by 40% per task and increasing output speed by 30% compared to Opus 5. Cache reads now cost $0.20 per 1 million tokens, a 60% reduction, and the model maintains a 1 million token context window.

The release comes as Anthropic prepares for an initial public offering and marks the first time an Opus model falls back to a less capable version for flagged requests related to cyber, bio, or frontier LLM development. Integrations are currently live through AWS, Google Cloud, Microsoft Azure, GitHub Copilot, Cursor and Devin. Subscription users on Pro, Max, and Team plans receive a 20% increase in five hour session limits and a one-time rate limit reset, with Sonnet 5.5 and Haiku 5.5 arriving in the coming weeks.