An independent, unofficial Codex reset signal.
Codex Reset Radar is a small, independent website that watches public signals about ChatGPT Codex usage limits and tells you, roughly, when the next reset window opens. It is not affiliated with OpenAI. It does not connect to your account, and it does not read any private data.
What this site is
Codex Reset Radar tracks publicly observable Codex usage-limit signals and surfaces them on three pages: a live snapshot, a history feed, and a statistics view. The goal is a glance-friendly answer to a recurring question: when does my Codex quota reset?
The site is maintained as a personal project by an independent operator. There is no company behind it, no team, and no funding relationship with OpenAI or any other AI vendor.
What this site is not
- Not affiliated with, endorsed by, or partnered with OpenAI.
- Not a tool that connects to your ChatGPT account. We do not ask for, store, or process login credentials, session cookies, or any personal data.
- Not a source of authoritative reset times. OpenAI does not publish a public, machine-readable schedule for Codex usage windows, and we do not have access to one. Treat every number on this site as an estimate, not a promise.
How the signal is built
Each reset estimate combines three signal layers:
- Direct public reset signals — public OpenAI/Codex-related posts or status context that explicitly mention resetting Codex usage limits.
- Product-release context — launches, milestones, and Codex product updates, tracked as context unless they explicitly mention quota or rate-limit resets.
- Social chatter from public posts on X and Reddit, used as a sentiment cross-check, not as ground truth.
The full method, including what we filter out and why, is documented on the Methodology page. If a number looks wrong, it usually means one of those three layers moved.
Who maintains it
One independent maintainer, working in their own time. The maintainer is not named on the site on purpose: this is a utility, not a personal brand, and adding a byline would not make the data any more correct. Contact details for corrections and takedown requests are listed in the Contact section below.
Corrections and takedowns
If you spot a factual mistake, a wrong attribution, or content that should be removed, please reach out. We respond to legitimate correction requests on a best-effort basis. We do not promise instant turnaround, and we do not run a 24-hour help desk.
Contact
The fastest way to reach the maintainer is by email:
Use it for:
- Factual corrections on a published reset signal.
- Source-attribution fixes (wrong author, wrong quote, broken link).
- Takedown requests.
We do our best to reply within a few business days. We do not provide account help, billing help, or any kind of OpenAI support. If your ChatGPT or Codex account is broken, please contact OpenAI directly — we cannot help with that.
First published: 2026-09-09 · Last updated: 2026-09-17