Anthropic has launched its latest AI model, Claude Opus 5, approximately two months after the previous Opus upgrade. This new version aims to offer intelligence close to the frontier of Claude Fable 5 at half the cost.

Opus 5 represents an upgrade from version 4.8 and is described as a "thoughtful and proactive model." Fable 5, Anthropic's most capable model, was initially withheld due to cybersecurity concerns but has been released with safeguards. Opus 5 is now the default model on Claude Max and the strongest option on Claude Pro, designed for efficient daily use.

Alongside Opus 5, Anthropic introduced mid-conversation tool changes for developers on the Claude Platform, allowing dynamic tool selection without invalidating the prompt cache. Additionally, automatic fallbacks on the API enable requests flagged by safety classifiers on Opus 5 or Fable 5 to be rerouted to another model, ensuring requests are always processed.

Claude Opus 5 is available immediately across all platforms. The pricing remains consistent with Opus 4.8, at $5/million input tokens and $25/million output tokens. Anthropic is also offering free access to premium Claude features for teachers.