Anthropic has pulled back from a controversial pricing update that could have dramatically increased costs for developers using its Claude Agent SDK. The company decided to pause these plans just as they were due to take effect, allowing users to continue enjoying their current generous usage limits.

The initial plan was to treat the use of Claude Agent SDK separately from standard Claude usage, applying API rates and capping it by subscription tier. This shift would have meant that developers might pay more for the same amount of usage, or less if they could generate enough API calls within their monthly cap.

However, Anthropic’s sudden reversal now means that users can continue to use the SDK without these additional costs. This decision comes as a relief to many third-party apps and power users who would have been hit hard by the change. The move highlights the ongoing debate around AI pricing and accessibility in the tech industry.

One analyst notes that Claude Opus users could start saving from their subscription after just two to three messages per day, potentially turning a costly service into one of significant value through API usage alone. This suggests that Anthropic’s change was not only about revenue but also user experience and market perception.