MyJackfruit

Help
Browse topics · Development

Development

← Development

My AI says I have reached my usage limit. What now?

Short answer

Wait for the limit to reset, or add usage credits or a bigger plan from your AI's maker; what your AI saved stays.

Deeper

The limit is your AI maker's, not GilbPlatform's: for your AI you pay its maker, Anthropic, OpenAI or xAI. GilbApp has no setting of its own for what your AI may spend.

Claude. A Claude plan has usage limits, on most plans every 5 hours and every week. Anthropic says the 5-hour period "starts when you send your first message, and your limit resets in full five hours later", and Weekly limits reset at a fixed time each week that is assigned to your account. Settings > Usage at claude.ai shows the time remaining and your next weekly reset. Anthropic gives three ways on: you can wait for it to reset or upgrade your plan; If you upgrade, your new plan's limits apply right away; and "With usage credits enabled, your work continues on usage credits instead of stopping." Usage credits are Anthropic's pay-as-you-go use beyond a plan's limits, charged separately at its standard rates; the answer What will my AI cost while I plan and build, and can I set a limit? shows how to turn them on and cap them (https://support.claude.com/en/articles/11647753-understanding-usage-and-length-limits and https://support.claude.com/en/articles/12429409-extra-usage-for-paid-claude-plans, read 13 November 2026).

In Claude Code in Terminal, the window where you type commands. Anthropic says that when a limit stops it mid-task, Claude Code "waits in the open session and continues the task on its own after the limit resets", with lines such as Usage limit reached · limit resets and the time. Keep that window open. Claude Code waits like this on its own only when you are signed in with a Claude subscription, it is version 2.1.234 or later, and the reset is less than 24 hours away. For a weekly limit that resets later, it opens a menu once; pick Wait here, then continue automatically to start the wait. It does not send your last message again, and it still asks you before acts that need your yes, so look in now and then. After three such waits in a row it stops; type /rate-limit-options and press Return to choose what to do. If the message names one model, such as You've hit your Opus limit, changing with /model to a model of another kind, the version of Claude you work with, lets you go on (https://code.claude.com/docs/en/interactive-mode, https://code.claude.com/docs/en/costs, read 13 November 2026).

In GilbApp's chat with Claude Code, the chat uses the Claude Code on your computer, so the same plan limits apply. If an error line shows above the box, your message did not get an answer: send it again after the reset.

Codex, OpenAI's AI for writing code. OpenAI counts use per 5 hours, and says Weekly limits may also apply. Its usage page, https://chatgpt.com/codex/settings/usage, shows your limits and when they reset, and in Codex /status shows what is left. OpenAI says Plus and Pro users can purchase additional credits to continue working. OpenAI also offers paying with an API key, a secret code for paying by use, but GilbPlatform's Codex steps say: Hand-offs run only on your ChatGPT sign-in, never on an API key. (https://learn.chatgpt.com/docs/pricing, read 13 November 2026).

Grok. xAI sets SuperGrok's limits on its own pages; we could not read them. When xAI answers GilbApp's chat that it has too many requests just then, GilbPlatform sends the same request again, at most twice in a row, after about 5 and then 20 seconds, or after the wait xAI asks for, up to 30 seconds; if that fails too, the chat shows an error line.

Your work is not lost. What lasts is what is saved in your vaults, and the files your AI wrote on your computer stay there. When you go on, Kai-Zen reads the project's setup, its specs and its last round of work, and does not ask again what you already answered; the answer I stopped in the middle. Where do I continue? shows how. You can also go on with another AI you have connected, on its own subscription; the answer Can I switch to another AI later, or use more than one? shows how. To see what used so much, see How do I see what my AI use has cost so far, and why was one day so much?

Anthropic, OpenAI and xAI set their plans, limits and prices, and may change them.

This answer is written for a computer screen.

Source: Anthropic, read 13 November 2026: https://support.claude.com/en/articles/11647753-understanding-usage-and-length-limits (five-hour period; weekly reset; wait, upgrade, usage credits; Settings > Usage); https://code.claude.com/docs/en/interactive-mode, section Wait for a usage limit to reset (waits and continues on its own, v2.1.234 or later; Usage limit reached · limit resets; does not resend the last message; still asks for permissions; stops after three automatic continuations; /rate-limit-options); https://code.claude.com/docs/en/costs (You've hit your Opus limit, switching with /model). OpenAI, read 13 November 2026: https://learn.chatgpt.com/docs/pricing (five-hour estimates; Weekly limits may also apply; Plus and Pro can purchase credits; API key; usage page; /status). xAI: no page could be read. GilbPlatform 0fc5a548 (https://id.gilb.com/version, read 13 November 2026): src/ai/complete-chat.ts:155-163 (429, 500, 502, 503, 504 sent again; waits 5 s then 20 s; Retry-After up to 30 s). GilbApp 49b1ae53 (https://gilb.com/app/version.json, read 13 November 2026): src/components/agent/AgentChat.tsx:821-822 (a failed turn shows an error line). Live help pages, sentences shared: What will my AI cost while I plan and build, and can I set a limit? (limits every 5 hours and every week; usage credits; no GilbApp spending setting); What do I need to get started? (Codex hand-offs only on the ChatGPT sign-in); What can my AI see and do in my vault, and does it remember? (what lasts is saved in your vaults); I stopped in the middle. Where do I continue?; Kai-Zen says something I do not understand, or stopped. What now? (error line: send it again); Can I switch to another AI later, or use more than one?

Checked against GilbApp d529620.