{"slug": "claude-code-usage-limit-reached-how-the-5-hour-and-weekly-limits-work", "title": "Claude Code usage limit reached: how the 5-hour and weekly limits work", "summary": "A developer measured Claude Code's usage limits on a single Pro/Max account, finding that 1% of the 5-hour session limit corresponds to roughly 166k new tokens but about 11.2 million cached tokens, based on 61 observations over two days. The analysis attributes window exhaustion mainly to new-token activity — subagents (about 2% of a 5-hour window each), re-sending context after the prompt cache expires, large tool output and file reads, and large starting contexts — and the developer released a tool, limitcast, that forecasts whether a window will run out before it resets and surfaces the readings in the Claude Code status line.", "body_md": "\"You've hit your limit\" usually arrives in the middle of something. Here is how Claude Code's limits work on a Pro, Max or Team plan,\n\nwhat actually moves them, and how to see the end of a window coming instead of finding out from the error.\n\n**There are two limits.** A session limit that resets every five hours, and a weekly limit that resets at a fixed time each week.\n\nWhat fills them is mostly **new** tokens: on my account, 1% of the 5-hour limit was about 166k new tokens but about 11 million\n\ncached ones.\n\n-\n\n**The 5-hour session limit.** Anthropic's help center: paid plans have \"a session limit that resets every five hours, and weekly\n\nlimits that reset once every week\"\n\n([How do usage and length limits work?](https://support.claude.com/en/articles/11647753-how-do-usage-and-length-limits-work)).\n\n-\n\n**The weekly limit.** It \"reset[s] at a fixed time each week that is assigned to your account\", whenever you started using Claude\n\n([What is the Pro plan?](https://support.claude.com/en/articles/8325606-what-is-the-pro-plan)).\n\n-\n\n**Where to see them.** Settings → Usage on claude.ai shows both as progress bars, with the time left in the current session\n\n([Usage limit best practices](https://support.claude.com/en/articles/9797557-usage-limit-best-practices)). Claude Code also\n\npasses the same readings to your status line, which is how a tool can show them while you work.\n\nWhen either one is used up you wait for its reset, or keep going on paid extra usage if you've turned that on. A full weekly limit\n\nstops you even with room left in the 5-hour window.\n\nThe meter is a percentage, and Anthropic doesn't say how many tokens a percent is. I measured it on my own account by comparing each\n\nrise of the 5-hour reading with the tokens my sessions used in between (61 observations, two days; the method is in\n\n[How many tokens is 1% of your Claude Code limit?](https://limitcast.dev/blog/tokens-per-percent/)):\n\n|  | Tokens per 1% of the 5-hour limit | \n|---|---|\n| New tokens (input, cache writes, output) | ≈ 166k | \n| Cached tokens (cache reads) | ≈ 11.2M | \n\nOne account, one plan. A fit on the previous day gave about 200k and 10.7M, so read these as a range.\n\nA cached token counted roughly 1/67 as much as a new one. Claude Code re-sends your whole conversation on every model call, but most\n\nof it comes back from the cache, so a long session is cheaper than its size suggests. What runs a window out is anything that sends\n\n**new** tokens:\n\n**Subagents.** Each one writes its own context from scratch. A typical one cost me about 2% of a 5-hour window ([Do subagents burn your limit faster?](https://limitcast.dev/blog/subagents/)).\n\n**Coming back after a long pause.** Once the prompt cache expires, the next message re-sends the whole context as new tokens ([Why a long idle pause costs you](https://limitcast.dev/blog/prompt-cache-idle/)).\n\n**Big tool output and file reads.** Everything new that enters the conversation is new tokens once.\n\n**A large starting context.** Instructions, tool lists and skills are written once per session and every subagent, then re-read on every call.\n\nThe percentage alone doesn't say whether you'll make it to the reset. That depends on your pace: 70% with an hour left is fine at a\n\nslow pace and a lockout in twenty minutes during a burst of subagents. limitcast does that arithmetic for you:\n\n``` bash\n$ limitcast status\nwindow used burn runs out resets today status\nclaude 5h 71% +18.2%/h 16:40 17:00 - AT_RISK\nclaude 7d 41% +0.6%/h - Sun 03:00 6% left OK\n```\n\nThe 5-hour window is forecast from the last hour's pace; **AT_RISK** means it runs out before it resets.\n\nThe weekly window uses your pace over the last full day, nights included, and shows today's even share of the week, so one busy hour doesn't cry wolf.\n\nThe same line sits in your Claude Code status line, from the readings Claude Code already passes to it. No login, nothing leaves your machine.\n\nThe free version gives you the forecast and the status line. [Pro](https://limitcast.dev/#pricing) also stops new subagents when a window is on\n\npace to run out within 30 minutes, moves them to Sonnet while it's tight, and sends a desktop notification before you hit the limit.\n\n**Don't start a fan-out near the end of a window.** Ten subagents at about 2% each are a fifth of it.\n\n**Compact long sessions** before a pause, not after: `/compact` shrinks what the next call has to re-send.\n\n**Queue the heavy job for after the reset** instead of starting it with 10% left.\n\n**Check what loads before you type.** Unused MCP servers and skills are re-sent on every call and to every subagent: `limitcast trim` measures them.\n\n```\nnpm install -g limitcast\nlimitcast install # adds the forecast to your status line; backs up settings.json\nlimitcast status # every window: used, burn rate, runout time, reset\n```\n\nNode 20+, macOS or Linux. [Install details](https://limitcast.dev/#install) · `limitcast uninstall` restores your settings.\n\nlimitcast forecasts when each Claude Code and Codex window runs out, and its Pro plan stops runaway subagents before they burn the rest of it.\n\n[See what limitcast does](https://limitcast.dev/) · [Pricing](https://limitcast.dev/#pricing) ·\n\n[How many tokens is 1%?](https://limitcast.dev/blog/tokens-per-percent/) · [Subagents and your limit](https://limitcast.dev/blog/subagents/) ·\n\n[Idle pauses and the prompt cache](https://limitcast.dev/blog/prompt-cache-idle/)", "url": "https://wpnews.pro/news/claude-code-usage-limit-reached-how-the-5-hour-and-weekly-limits-work", "canonical_source": "https://dev.to/samstull9007/claude-code-usage-limit-reached-how-the-5-hour-and-weekly-limits-work-3mn1", "published_at": "2026-10-26 03:38:33+00:00", "updated_at": "2026-10-26 03:42:48.733147+00:00", "lang": "en", "topics": ["ai-tools", "ai-products", "developer-tools", "large-language-models"], "entities": ["Anthropic", "Claude Code", "limitcast"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/claude-code-usage-limit-reached-how-the-5-hour-and-weekly-limits-work", "markdown": "https://wpnews.pro/news/claude-code-usage-limit-reached-how-the-5-hour-and-weekly-limits-work.md", "text": "https://wpnews.pro/news/claude-code-usage-limit-reached-how-the-5-hour-and-weekly-limits-work.txt", "jsonld": "https://wpnews.pro/news/claude-code-usage-limit-reached-how-the-5-hour-and-weekly-limits-work.jsonld"}}