Claude's new usage limits are insane.
Users are reporting that Anthropic's Claude Opus 4.8, particularly when combined with features like the 1 million token context window and UltraCode, is consuming usage limits at an unexpectedly rapid rate. One user noted that a single prompt used 21% of their 5-hour limit, attributing this to the model spawning numerous agents that independently process the large context window. While the underlying token math is understood, some users suspect this behavior is an intentional strategy by Anthropic to encourage users to burn through their limits faster and potentially upgrade their plans. AI
IMPACT Users are experiencing rapid consumption of usage limits with Claude Opus 4.8, potentially impacting workflow and subscription costs.