Is ChatGPT down — or is it just you?

Start with the official status page at status.openai.com, which lists current incidents, degraded performance and scheduled maintenance for ChatGPT and the API. As a cross-check you can look at third-party outage trackers like Downdetector, which show whether other users are reporting problems at the same time. If both look normal, the issue is more likely on your side.

The official page is status.openai.com. It is run by OpenAI and shows the live status of ChatGPT, the API and related services, plus a history of recent incidents and any planned maintenance. It is the most reliable single source for confirming whether ChatGPT itself is having problems.

Common signs of a real outage include the status page reporting an incident, errors that affect everyone (such as "at capacity", 500, 502 or 503), the same failure across different networks and devices, and a spike of reports on outage trackers. If only you are affected and the status page is green, it is more likely a local issue.

Try to reproduce the problem somewhere independent: a different network (for example mobile data instead of Wi-Fi), another device, or an incognito window. If ChatGPT works in any of those, the original problem is likely local. If it fails everywhere and the status page reports an incident, it is probably an outage on OpenAI's side.

Outage trackers such as Downdetector are useful as a cross-check because they show whether many people are reporting problems at once, but they are based on user reports rather than OpenAI's own systems and can lag or over-report. Treat the official status.openai.com as the primary source and use trackers as a secondary signal.

If the status page confirms an outage, there is nothing to fix on your side, so avoid hammering refresh. Wait for OpenAI to recover, and in the meantime keep working in another model. A multi-model workspace like MultipleChat lets you switch to Claude, Gemini or another model for free so you don't lose time during the outage.

It varies. Many incidents are short-lived and resolve within minutes once OpenAI mitigates the cause, while larger outages can take longer. The status page is the best place to follow progress, as OpenAI typically posts updates as they investigate and resolve an incident. We avoid quoting fixed durations because they differ from one incident to the next.

If status.openai.com shows no incident but you still get errors, the problem is most likely local. Try a refresh, an incognito window, a different network, turning off your VPN or proxy, clearing ChatGPT cookies and cache, or disabling browser extensions. A blocked IP can also cause an access error (such as error 1020) even when ChatGPT itself is healthy.

Usually no. An outage affects many users at once and is unrelated to your individual account. Errors during an outage are temporary server problems, not bans. If you still cannot access your account after an incident is resolved and after trying the standard local fixes, you can contact OpenAI support.