ChatGPT Not Working? Tell an Outage From a Problem on Your Side

When ChatGPT ceases to respond you first have to determine whether the issue lies with you or OpenAI. It is best to resolve this before deleting your data or disconnecting your VPN. Depending on what corrective action you take, you may lose your privacy or have been logged out and none of the solutions you have applied will assist with genuine failures – such as real time outages. According to the troubleshooting guide for ChatGPT errors in the troubleshooting section of OpenAI’s support page, problems with ChatGPT can arise from your internet connection, browser extension(s), and/or temporary server issues.

If ChatGPT is not working properly there are three areas that you can check for problems: ChatGPT itself, your internet connection, and your browser session. You can complete three quick checks without making any changes to determine which area is having issues: the OpenAI Status Page; a Private browsing window; and a second internet provider. While none of these will allow you to identify the exact cause of the issue, each one will help you determine where to focus your time in the troubleshooting sections of this guide.

The 3-Minute Fault Isolation Test (Decision Tree)

The “three minutes” is a goal, and therefore not guaranteed. You are encouraged to reach your goal within three minutes; however, if it takes longer because you had to use mobile data or needed a second phone to complete your sign up, you still have all the time you need. Don’t let that discourage you from going through the whole process! You’ll keep all your chat history, settings, and security tools; you just need to split things out into service, connection and session, which is how this site works, based on info from OpenAI help pages. OpenAI never calls it that!

Step 1: Check the status page for the part that’s failing. Open OpenAI Status and look at the ChatGPT group of components, not just the headline banner. If there’s an active incident that matches what you’re seeing, stop here. Wait, and watch the incident for updates. Changing things on your own device won’t fix a problem on OpenAI’s side. If the page is green, go to Step 2. A green page doesn’t clear OpenAI, for reasons explained in the next section.

Step 2: Open a private window and start a new, short chat. OpenAI’s error guide recommends using an incognito or private browsing window. Its guide to slow responses adds that incognito mode disables most extensions by default. Go straight to chatgpt.com, not a link from somewhere else. Sign in with the same method you used when you signed up, whether that’s Google, Microsoft, Apple or an email and password. Then send one plain sentence.

• If it works in the private window, the fault most likely sits in your normal browser: an extension, cookies or stored site data. Go to the “Something Went Wrong” section.
• If it fails there too, go to Step 3.

Step 3: Switch to a different network. If you’re on Wi-Fi, turn it off and use mobile data. If you’re already on mobile data, try a Wi-Fi network you trust. OpenAI’s login guidance gives switching from corporate Wi-Fi to mobile data as a test, and its error guide suggests a different network or device. One short test prompt uses very little data. Expect some slowness on mobile data, though: OpenAI notes that it may have lower bandwidth, so slower doesn’t mean broken.

• If it works on the other network, the problem is most likely in the path your first network takes: the router, a VPN, a web filter or a firewall. Go to the “Network Error” section.
• If it fails on both networks, and also on a second phone or laptop, the service, your account or one feature becomes the more likely cause. Go to the next section, and then to Support.

One of the community members at OpenAI explained how the Wi-Fi/cellular split helped identify their problem. After numerous attempts to connect without success, they stated, “I have verified that it only happens when I connect to Wi-Fi but with a data plan it works normally.” This may not represent everyone, but it does illustrate the separation that this test provides between networking problems and other possible causes.

Limit the issue to one individual conversation, model, or feature. If you are able to create a successful chat with new conversations, but unable to locate your previous conversations, the issue lies within those conversations. OpenAI recommends starting a new chat when the last chat is lengthy or has several previous chats, in addition to switching to a different model in a brand-new chat. Even if regular text-based chats work fine, you may have problems with uploading files, generating voice or images. Make sure to check OpenAI’s current status page regarding each feature.

How to Tell If ChatGPT Is Actually Down (Beyond the Status Page)

If you’re asking yourself whether “ChatGPT is down today”, you should first check the status page. However, the status page’s headline does not tell the full story. Below the OpenAI Status heading on the status page, OpenAI states that its availability data “is provided in an aggregate view across all changes, models, errors and tier levels.” OpenAI also states that “individual customer availability can differ based on subscriptions, individual models and API features.” In other words, if the headline is green, it indicates that OpenAI is functioning normally overall. However, this does not mean that your subscription level, your location or the specific feature you are using, all work correctly.

OpenAI maintains an incident log showing how limited an outage can be. All of the incidents below are resolved and listed only as past examples:

• On 04/09/2026, OpenAI reported that users in the APAC region may face increased errors in ChatGPT, Work, image generation, file upload, Voice and Codex Cloud. That was one region, while other regions weren’t listed as affected.
• At the end of August 2026, an incident covered elevated errors in ChatGPT conversations for Free and Go plans only. It was dated 01/09/2026 on the page as displayed.
• On 23/09/2026, the affected group was the reverse: elevated error rates for ChatGPT across Plus and Pro plans.
• On 20/08/2026, the problem was one feature. OpenAI investigated errors when using Thinking mode and Image Generation.

Therefore, your friend, who has access to a different subscription tier stating “it works for me!” does not imply that there is an outage with your subscription. When a new incident is posted, it is important to read which components of the system, the subscription level and location that is affected and the current status of the incident. The incident logs presented above progressed through different stages, i.e., “investigating”, “identified”, “monitoring” and “resolved”. The “monitoring” stage indicates that a fix has been implemented for an issue, but recovery time is still being assessed.

There are times when ChatGPT is not functioning properly, even when its page is indicating it should be functioning. On September 5, 2024, a user on the OpenAI Community stated that they were not receiving any response from ChatGPT, but there was nothing about it being offline on the Status page. Several users on the same thread also experienced an inability to receive responses from ChatGPT around the same time as this individual. The thread clearly indicated that multiple individuals were unable to access the system, however, there was no explanation as to why this issue occurred, nor does it provide proof that the status page always provides timely updates.

Crowdsourcing sites provide validation of the issues being experienced, however, they do not take the place of the official status page. Downdetector maintains a dedicated Nigeria page for OpenAI that is created based on user-submitted reports of issues. A spike in activity on this page indicates there are other users experiencing problems with ChatGPT. This activity does not guarantee an occurrence of an event, nor does a lack of activity on the page prove that there is nothing wrong with your environment. For these reasons, your best options remain conducting your tests within a private browser window and with a separate internet connection when there is no response from the status page.

Fixing the ChatGPT “Network Error” Loop (VPNs and WebSocket Drops)

A ChatGPT Network error does not indicate a total failure of the internet connection. In general, you will see a fully loaded page, but you may see the reply cut off in the middle of sending. In the OpenAI error documentation, it states that the two messages “Network Error Occurred” or “Error Occurred While Connecting to Websocket” indicate that there is an unstable connection between your device and OpenAI’s servers. The important part of this message is “Unstable”. Your device is unable to establish a reliable connection.

This is why an ordinary page load can succeed while a reply fails. According to OpenAI’s network recommendations, some ChatGPT features use a secure WebSocket (a connection that stays open while the reply streams in) in addition to ordinary web requests. For ChatGPT, that connection runs to wss://ws.chatgpt.com over TCP port 443. OpenAI warns that if WebSocket traffic is blocked, features “may stall, disconnect, or fail to stream updates correctly.” Web filters, proxies and TLS inspection can also block the connection or close it early, and so can idle timeouts, which cut a connection that looks inactive.

The table below matches when the error shows up to what it most likely points to. Treat each row as a lead to check, not a diagnosis.

In short, the error on its own tells you little. When it happens and what’s switched on at the time tell you much more.

When switching tabs kills the reply

To save memory, desktop Chrome deactivates tabs you aren’t using through its Memory Saver feature. When you go back to a deactivated tab, it reloads. Supercharge Browser, a company that sells a tab-management extension, says a mid-reply error after you switch tabs comes from Chrome discarding the tab and killing the WebSocket. That’s the company’s own explanation, not a finding from OpenAI. The fix doesn’t need anyone’s extension, though. Google’s instructions are: in Chrome, select More, then Settings, then Performance, and next to “Always keep these sites active”, select Add and enter chatgpt.com. Google also notes that you can’t change these performance settings on Chrome for Android or iOS, so this fix only applies on a computer.

Not every problem is caused by tab sleep. Some users may never find out what’s really causing their issue. A Plus member of the OpenAI Community posted in December 2023 about a network problem that happened to them repeatedly: “I receive the error message, Network Error, every other post!” They asked if it was a widespread problem. Although other members of this thread shared other theories, they were unable to pinpoint a cause for this type of error. If you have experienced ongoing network issues, first check all items listed below.

VPNs, WARP and security apps

The initial recommendations from OpenAI regarding this issue are to [disabling VPNs or proxy connections and turning off Web Protect or other security filters]. Use this as a temporary trial only, do not rely on it as a permanent solution. Furthermore, do not use this method while using public Wi-Fi unless you use the VPN to ensure the privacy of your information. Cloudflare’s WARP app has modes where all traffic from your device goes through WARP so that all requests made by ChatGPT from your device will be made through WARP.

One Cloudflare Community user reported that with WARP+ on, about one in three requests failed. They fixed their own case by excluding OpenAI domains from the tunnel. That’s one person’s result on one Mac. Cloudflare’s own documentation notes that traffic you exclude this way bypasses the client’s filtering, so check the current steps for your own device before changing any split-tunnel settings.

A VPN may also create another type of issue. OpenAI associates its “suspicious activity” warning with users that are utilizing VPNs or proxies that route traffic through flagged IPs and therefore are routing their requests through IP addresses that have been marked as suspicious. While Nigeria is one of the countries where ChatGPT is supported, that same webpage advises that if you are utilizing ChatGPT from an unsupported country your account may be blocked or suspended. If your VPN is routing your requests through an unsupported country, it is recommended that you stop doing this and fix the issue immediately.

Do not attempt to bypass an organization’s controls while using the organization’s managed network, such as a school or office. Send the OpenAI Network Guidance to your IT representative; the guidance was created for IT teams to assist IT professionals in how to allow WebSocket traffic through port 443, perform an “Upgrade: WebSocket” handshake, and set up TLS inspection rules and idle timeout settings.

Why ChatGPT Says “Something Went Wrong” (And How to Clear It)

OpenAI’s statement “Something went wrong” is an overall message that indicates either a momentary server fault or a problem with your local device. The message by itself does not inform you which type of issue you have. The result of your isolation test should indicate whether the type of problem is temporary in nature. Prior to performing the isolation test, copy any prompt you input but did not send into a notes application as it is possible that you could lose your unsent prompt if you refresh the webpage.

Work through the fixes from least to most disruptive:

1. Refresh the page or start a new chat. If the error only happened once, this is often enough, and OpenAI lists it first.
2. Check OpenAI Status for the ChatGPT component you’re using.
3. Try a private window if you didn’t in the isolation test. If it works there, your normal browser setup is the problem.
4. Turn off extensions one at a time. OpenAI singles out privacy and security tools in particular.
5. Clear ChatGPT’s cookies and site data. This will sign you out of ChatGPT in that browser. Make sure you know which sign-in method you used before you do it. OpenAI’s slow-response guide specifically says to clear the site data, not just the cache.
6. Turn off your VPN or secure DNS tool for a test, or switch networks, as covered in the section above.
7. Try a smaller request. Use a new chat with plain text and no attachment. If that works, the size of your first request or a particular tool may be involved, but that’s a clue, not proof.

If you get the “Something went wrong” message while trying to log into your account, then please refer to OpenAI’s login-specific steps related to Logging into OpenAI. If you have a generic sign-in error from OpenAI, you should wait at least sixty seconds before trying to sign back in to see if the error resolves itself. OpenAI suggests that you check for updates to the Google Play Store application if you receive a “Something went Wrong” error on your device running the Android operating system, which can happen when your Google Play Store version is out of date. To perform this check, go to the Google Play Store and tap on the profile image icon and follow the series of options from Settings > About > Play Store Version.

Clearing site data won’t delete your chats

Clearing site data does not remove your ChatGPT conversation history. According to OpenAI’s webpage dedicated to managing your ChatGPT chats, any and all chats you have created (both current and archived) will stay associated with your account until you delete them or your organization has a retention policy governing when you must delete them. Clearing cookies from your web browser is a differing process than deleting Chats from ChatGPT, therefore, when you log back into ChatGPT you should still have access to all your previous conversation history.

What you really need to watch for is which account you sign back into. OpenAI warns that different sign-in methods can create separate accounts and that it doesn’t merge them automatically. If your sidebar looks empty after signing back in, check that you used the same Google, Apple, Microsoft or email login as before. And never use Delete all chats as a fix. OpenAI states that this action cannot be undone.

What to Do When ChatGPT Is Not Responding or Stuck Generating

If ChatGPT returns a slow response, that’s different than if it provides no response. If ChatGPT stops responding and keeps spinning with no new text, you should follow the specific order of steps described in OpenAI’s error guide. The first item is to copy any part of the response that is already visible on your screen. On 4/8/2026, OpenAI said that in some situations, parts of the conversation might not be saved. Therefore, you should not assume that anything you wrote that was only partially completed will remain available when you reload the page.

1. Wait 30–60 seconds to see whether the reply finishes.
2. Select “Stop generating”, then “Regenerate” to retry the same request.
3. Start a new chat and send your prompt again.
4. Hard refresh the page with Ctrl + Shift + R, or Cmd + Shift + R on a Mac. In the desktop app, reload the app instead.
5. Sign out and back in, using the same sign-in method as before.
6. Open a private window with extensions disabled, then try turning off your VPN, proxy or secure DNS tool.
7. Try another browser, device or network, and check the status page again.

Stopping ChatGPT is safe. In some cases, stopping ChatGPT is the only solution. An OpenAI Community member described a bug related to a specific ChatGPT model on 30/8/2026. In this user’s case, their long ChatGPT reply was visibly finished while the page still acted as though it was generating, but their browser was still displaying the loading icon. They also stated that clicking the Stop button finalizes the response immediately. This user’s issue occurred in Incognito Mode and after clearing their browser’s site data. Therefore, not all freezes can be resolved by users.

If ChatGPT doesn’t appear to load at all (e.g., an empty browser window), and you’re stuck at the spinning icon and no error message appears, OpenAI’s recommended next steps are as follows: perform a hard refresh, sign out, then sign back in to your account, clear the site data in your browser, disable any content blockers, and try using a different browser or setting up a new profile.

Taking it to OpenAI Support

If you experience the problem on all browsers, devices, and networks, you should contact Support for assistance. When you contact Support, you need to provide them with a HAR file (the HAR file contains a record of network traffic from your web browser), any console errors with dates/times, the name of the model you used, and the link to or ID of the conversation that you had with the model. You can access Support via the chat bubble at help.openai.com (the chat bubble is located in the bottom right corner of this page), and you can find instructions on how to create a HAR file in your browser’s developer tools on the same page as well. Make sure to include the timezone in which the timestamps of the console errors were recorded. For Nigeria, that is WAT (West Africa Time).

When you save the HAR file, you should be aware that OpenAI states that all or a portion of the HAR files may contain personal and/or sensitive information and that many browsers such as Google Chrome will automatically erase all of the information in the HAR file when it is saved. Before sending the HAR file, review its contents carefully; do not post the file publicly and do not share your password or any one-time codes. Additionally, note that the ChatGPT desktop app does not have any developer tools for generating a HAR file, so you must recreate the issue using a web browser to obtain a HAR file for your Support request.

What to Check First

To determine whether there is an issue with OpenAI’s status pages, begin by reviewing both the overall status as well as the specific components that are impacted by your own issue. Next, check for issues using both a private browser session (which should isolate you from most browsers’ cached data) and a second location/ISP. Comparing these two ways will allow you to know if you need to either wait on OpenAI to resolve an issue or if there is something wrong with your network connection (and thus should be repaired) or if you need to delete the cached data from your browser or disable your VPN. You will want to resolve those issues after testing has determined which of those types of solutions needs to be performed to resolve the problem you are having with OpenAI. Once everything has been resolved, then you can go back to utilizing your privacy protection tools when using OpenAI services.

Frequently Asked Questions

Will clearing my browser data delete my ChatGPT chats?

According to OpenAI, any chats you have with ChatGPT are saved to your account until you delete them. Clearing the browser history of your site does not delete your saved chats; it only removes the cookie-based login device from that particular browser (you need to log back into OpenAI using the same method you had before; otherwise you may end up in a different account).

Why does ChatGPT work on mobile data but fail on my Wi-Fi?

That points to something in your Wi-Fi’s path, such as the router, a web filter, a VPN or a firewall. OpenAI notes that proxies and filters can block or close the connection that ChatGPT uses to stream replies. Ask whoever manages that network to check it.

Should I leave my VPN off for ChatGPT?

You should only turn off the VPN momentarily, purely as a test, to see if that loosens up the connection enough to allow ChatGPT to function. If so, you have a variable for further investigation. However, be sure to enable your VPN when finished testing. Also ensure the VPN is not routing you through a geographical location that does/do not correspond with the list of countries OpenAI supports.

Is it safe to send a HAR file to OpenAI Support?

It is possible, with caution, that HAR files could be created via ChatGPT. OpenAI suggests that HAR files may include personal or other sensitive data or contain personally identifiable or sensitive information but have noticed that in some cases (i.e.., using Chrome), the browser cleans the file before saving it. If you create an HAR file, send it via the official Support Chat only. You should never send your password or one-time code through the HAR.