Home / usage limits
Claude Code, Codex and Antigravity usage limits: how to see them
All three tools count your usage in windows. You get a 5-hour window, a weekly window, or both. Here is how each one works and where to check it.
The two windows
There is a 5-hour window and a weekly window. Which of the two you have depends on the tool and on your plan.
Claude Code
- Pro and Max plans have a five-hour session limit and a weekly limit.
- Ultra plans have a weekly limit. Verified Ultra accounts don't have a five-hour session limit.
- Max and Ultra also have a separate weekly limit for Fable.
- The limits are shared across Claude and Claude Code. Activity in both counts against the same limits.
- Anthropic does not publish a fixed message count. The limits are not a fixed number of messages.
Run /status in Claude Code to check your remaining allocation. Claude Code also shows warnings as you approach a limit.
If you hit the five-hour limit mid-response, Claude Code may finish the step in progress. You will see "Usage limit reached · wrapping up". That does not cover a new message.
Codex
- Plus and Standard Business usage is measured in five-hour windows. Pro plans currently have no five-hour limit.
- Local messages and cloud chats share the plan's allowance. Weekly limits may also apply.
- The per-model message estimates are not fixed message limits.
In Codex CLI, /status during a session shows your remaining limits. OpenAI also has a usage dashboard with limits and reset times.
Antigravity
The Antigravity CLI shows its quotas with /usage. Quotas are grouped by model family, with 5-hour and weekly windows. Some models, depending on the plan, have only a weekly window.
At a glance
| Tool | 5-hour window | Weekly window | How to check |
|---|---|---|---|
| Claude Code | Pro and Max. Not on verified Ultra. | Yes | /status |
| Codex | Plus and Standard Business. Not on Pro. | May apply | /status |
| Antigravity | Yes, for some models | Yes. Only this for some models. | /usage |
Plans change often. The official pages are the source of truth: Anthropic's Claude Code plan article and OpenAI's Codex pricing page.
Why three tools are hard to watch
Each tool has its own command, and each command lives inside its own terminal. To know where you stand, you stop what you are doing, type the command, read the answer, and switch to the next tool.
A quick check in one tool tells you nothing about the other two. If you run all three side by side, you end up checking all the time or not at all.
One top bar for all of them
Orbiterm puts each AI's 5-hour and weekly limits side by side in one top bar, as the tools themselves report them. The bar is always on top, so you can glance at it while you work.
It reads the AI tools' own local logs, read only. Prompts, code and file names stay on your machine. Besides the limits, Orbiterm shows every session as a planet on an orbit map, with each session's prompt, last tool call, context bar and spend.
FAQ
How many messages do I get in a 5-hour window?
There is no fixed number. Anthropic does not publish a fixed message count for Claude Code, and the limits are not a fixed number of messages. OpenAI also says the per-model Codex message estimates are not fixed message limits.
Do Claude and Claude Code share the same limit?
Yes. The limits are shared across Claude and Claude Code, so activity in both counts against the same limits.
What happens when I hit the 5-hour limit?
You can wait for the reset, enable usage credits, use a Claude Console account with API credits, or upgrade the plan. If the limit hits mid-response, Claude Code may finish the step in progress, but that does not cover a new message.
Get Orbiterm — free for macOS and Windows
Free desktop app for macOS 13+ (Apple Silicon and Intel) and Windows 10/11 64-bit. No sign-up, no paid tiers.