Home / Guides / Codex Usage
How to Check Your Codex Usage and Rate Limits
OpenAI Codex enforces usage limits on a rolling 5-hour window. Here's how to monitor your token consumption, understand your rate limits, and avoid getting locked out mid-session.
Last updated March 2026 · By Soren Starck
Understanding Codex Usage Limits
If you use OpenAI Codex for coding, you're subject to rate limits that determine how much you can use the tool within a given period. These limits are enforced on a rolling 5-hour window, meaning the tokens you used 5 hours ago gradually become available again.
The problem? There's no built-in warning system. You're coding, in flow, making progress — and suddenly Codex stops responding. You've hit your rate limit. Now you're stuck waiting, with no clear indication of when you can resume.
Method 1: Check Usage via the OpenAI Dashboard
The manual approach is to check your usage through the OpenAI platform:
- Go to platform.openai.com
- Log in to your account
- Navigate to Usage in the sidebar
- Review your token consumption and costs
Limitations: The dashboard shows historical usage data, but it doesn't show your real-time position in the 5-hour rolling window. You can see how many tokens you've used, but not how close you are to hitting your rate limit right now or when it resets.
Method 2: Monitor in Real-Time with SessionWatcher
SessionWatcher for Codex is a native macOS menu bar app that solves the monitoring problem completely. Instead of switching to a browser, logging in, and navigating dashboards, you just glance up at your menu bar.
SessionWatcher tracks:
- Real-time token usage — see exactly how many tokens you've consumed in the current window
- 5-hour window countdown — know exactly when your rate limit resets
- API cost tracking — monitor what you're spending in real-time
- Rate limit status — see how close you are to hitting the cap
- macOS notifications — get warned before you hit the limit
Skip the manual checks entirely.
SessionWatcher for Codex sits in your menu bar and shows everything you need — always visible, zero configuration. $1.99 one-time, no subscription.
How the Codex 5-Hour Rolling Window Works
The 5-hour window is a rolling limit, not a fixed one. Here's what that means in practice:
- At 9:00 AM, you start coding and use 50% of your token limit
- At 11:00 AM, you use another 40% — you're now at 90%
- At 11:30 AM, you use 10% more — you're locked out
- At 2:00 PM (5 hours after 9 AM), the tokens from your 9 AM session start freeing up
- Your available limit gradually increases as older usage rolls off
This rolling nature makes it particularly hard to track manually. SessionWatcher handles this calculation for you in real-time, showing both your current usage and when capacity frees up.
SessionWatcher vs. Manual Monitoring
| Feature | SessionWatcher | OpenAI Dashboard |
|---|---|---|
| Real-time token count | Yes | Delayed |
| 5-hour window tracking | Yes | No |
| Rate limit warnings | macOS notifications | None |
| Setup required | 10 seconds | Login each time |
| Workflow interruption | None (menu bar) | Switch to browser |
| Cost | $1.99 one-time | Free |
Frequently Asked Questions
How do I check my Codex usage?
You can check via the OpenAI dashboard at platform.openai.com/usage for historical data. For real-time monitoring including the 5-hour window, use SessionWatcher for Codex.
What is the Codex 5-hour rate limit window?
Codex enforces usage limits on a rolling 5-hour window. Tokens you used 5 hours ago start becoming available again. Without monitoring, you won't know when you're approaching the limit or when capacity frees up.
Why did my Codex stop working mid-session?
You've likely hit your rate limit for the current 5-hour window. Codex stops responding without a clear warning. SessionWatcher prevents this by alerting you before you reach the limit.
Is there a Codex monitor app for macOS?
Yes. SessionWatcher for Codex is a native macOS menu bar app that monitors your usage in real-time. $1.99 one-time, zero configuration, works in 10 seconds.