Claude Code usage limits: what counts, when they reset, and how to stop hitting them
If you use Claude Code on a Pro or Max subscription, you have probably seen the message that you have reached your limit halfway through a task. The limits are not hard to understand, but they are easy to lose track of, because they are shared, they reset on different schedules, and Claude Code burns through them faster than chat. The short version When you sign in to Claude Code with a…
Claude Code usage limits are shared across chat and coding sessions, resetting at different intervals depending on your subscription plan. On Pro or Max plans, you have a rolling 5-hour session and a weekly limit that resets weekly, shared with other Claude.ai models. API key users pay per token under their rate limits, without these shared limits.
During coding, agentic loops can cause one request to trigger many model calls, while large context and stronger models consume more of the limit per message. To extend your limits, break unrelated tasks into separate sessions and shrink long conversations using /compact to reduce context.
Plan around resets by scheduling large refactorings after the reset, rather than before. Use built-in tools like /usage to monitor your session and weekly usage, and consider open-source alternatives like ccusage, Claude Monitor, or Claude Monitor for more detailed insights into token usage, estimated cost, and reset times.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
