Claude Code usage limit reached: what to do

What does “You’ve hit your session limit” mean?

Your Claude subscription has a rolling usage allowance, and you used it up. Claude Code shows one of four messages: session limit, weekly limit, Opus limit or Sonnet limit, each with the time it resets. Until then it blocks further requests, according to Claude Code’s error reference.

Anthropic’s help center says Pro and Max plans have a five-hour session limit and a weekly limit, and that the limits are shared across Claude and Claude Code. A long chat in the Claude app and a long coding session draw from the same pot.

One detail saves time. The session and weekly limits cover every model, so switching models does not bring you back. The Opus and Sonnet limits are different: each applies to one model family, so /model to a model outside that family keeps you working.

What can you do right now?

Run /usage. It shows your plan limits and when they reset, so you know whether you are waiting twenty minutes or four days.

Then pick a path. The first three are Anthropic’s own options:

Wait. Since version 2.1.234, Claude Code can wait in the open session and continue the task by itself after the reset. It is on by default in interactive sessions signed in with a claude.ai subscription, and you need to keep the session open. It still asks for permissions as usual, so a task can stop on a prompt while you are away. The interactive mode docs cover how to cancel the wait.

Pay as you go. Anthropic lets paid plans continue past the limit at standard API rates, billed separately from the subscription, with a monthly spend limit you can set. Read their help article before turning it on.

Move up a plan. Anthropic suggests the next tier if you hit limits all the time. Prices and limits change, so check claude.com/pricing for current numbers.

Switch tools for the rest of the day. More on that below.

One more check. If an ANTHROPIC_API_KEY environment variable is set, Claude Code uses the key instead of your subscription, and you pay API rates instead of using your plan, per the same help center article. Check with test -n "$ANTHROPIC_API_KEY" && echo set, which prints only the word set and never the key.

Why did it run out so fast?

Usually because the conversation is long, not because the last prompt was big. Claude Code sends your whole conversation with every request, so a one-line question in a session that has been open all day still costs the whole history. Anthropic’s cost docs list the usual causes: long context, a cache miss after a break, subagents, scheduled tasks and compaction.

Parallel work is the other big one. The agent view docs say that running ten agents in parallel uses quota roughly ten times as fast as running one. Budget for that before you open a fourth terminal.

Heavy bursts hit two meters. Usage counts against the session and the weekly allowance at the same time, so one big run can drain the week before the five-hour window resets.

/usage has a breakdown that helps. On a paid plan it shows what share of recent usage came from skills, subagents, plugins and individual MCP servers, and it flags behaviors such as long context or cache misses when one is 10% or more.

How do you stretch the next window?

These five habits come straight from Anthropic’s cost guidance:

/clear between unrelated tasks, and /rename first if you want to come back. Use /compact with a short instruction, such as keeping test output and code changes. Use a smaller model for routine work and save the largest for hard problems. Turn off MCP servers you are not using. Write specific prompts: “add input validation to the login function in auth.ts” beats “improve this codebase.”

Claude Code also warns before you hit the wall, with a message such as “You’ve used 85% of your session limit.” You can show that continuously in a custom status line. Treat the warning as the moment to finish the current task and commit, not to start a new one.

How do you keep working with another agent?

The limit belongs to your Claude account, not to your project. Other agents have their own allowances, so the work can continue elsewhere. Codex usage is included in your ChatGPT plan, and OpenAI shows current limits in a usage dashboard and with /status in the CLI. Gemini CLI documents a daily request quota per sign-in type on its quotas page; as of October 2026 a personal Google account is listed at 1,000 requests per user per day.

A handoff is only as good as the brief. Before you switch, run git status and commit or stash what works. Then write five lines: the goal, what is done, what is left, the failing check, and the files that matter. Paste that into the new agent along with the test command, and ask it to run the check before it edits anything.

If you keep separate accounts of your own, such as one for work and one for personal projects, Anthropic documents running both: give each its own config directory with CLAUDE_CONFIG_DIR, for example alias claude-work='CLAUDE_CONFIG_DIR=~/.claude-work claude', as described in the authentication docs. Use accounts you are entitled to use, and follow each provider’s terms.

Switch agents without the copy and paste

SwarmPane’s Limit handoff is built for this moment. When a usage limit hits, you can switch to your other account, or from Claude Code to Codex and back, with a brief of the work that you edit first. Or you can resume after the reset, even if your computer slept.

SwarmPane runs the agent CLIs and accounts you already have, so the providers bill their own usage separately. Start with a 7-day trial for $1 and cancel before it ends if it is not for you.