ChatGPT Troubleshooting Manual: Unresponsive Messages, Model Unavailable, and Stuck Loading

If ChatGPT freezes, you can’t send messages, or you see an “unavailable” prompt, don’t rush to keep clicking Send. This article provides a ChatGPT troubleshooting checklist following the idea of “locate first, then fix,” helping you restore conversations and functionality quickly with minimal steps.

Start with three steps to pinpoint the cause: is it on your side or the server side

When troubleshooting ChatGPT errors, the first step is to switch to a different network environment: switch from Wi‑Fi to a mobile hotspot, or test the other way around once. The second step is to check the official status page for any service fluctuations (if the server is having issues, no matter how much you tinker locally, it’s hard to solve). The third step is to use the same account on a different device or in a different browser; you can quickly determine whether it’s a “device issue” or an “account/service issue.”

Messages won’t send, endless spinning: prioritize the browser and cache

For this kind of situation, the most effective troubleshooting combo is usually: refresh the page → log out and log back in → clear site cache and cookies. If you have ad blockers, script managers, or privacy-protection extensions installed, try disabling them temporarily, because they may block requests and cause the endless spinning.

If the problem occurs only in a specific browser, switching to an incognito window or another browser often restores it immediately. On mobile, prioritize updating the app version, and try again after turning off “battery saver/data saver mode” in the system settings.

Model unavailable or switching fails: don’t keep clicking—recreate the conversation first

When you see an incomplete model list loading, switching failures, or an “unavailable” notice, ChatGPT troubleshooting suggests creating a new chat first and then selecting the model again. Old chats—especially those with lots of content or many plugin/tool calls—can slow down during loading, making it look like “the model is broken.”

If it still doesn’t work, first confirm whether regional network instability is causing API timeouts; in that case, changing networks is more useful than repeatedly switching models. If the status page shows an incident, wait for recovery and avoid frequent clicking that could trigger short-term rate limits.

File upload/analysis fails: check size, format, and page permissions

For file-related ChatGPT troubleshooting, start with the three most common checks: whether the file is too large, whether the format is common, and whether the filename contains special characters. On the browser side, also confirm that the site permissions haven’t disabled “file access/downloads/pop-ups,” and whether your corporate intranet is blocking the upload domain.

If drag-and-drop upload doesn’t respond, using the “Choose file” button is usually more reliable; if the same file works in an incognito window, it’s most likely caused by cache or extensions—just clear them following the earlier ChatGPT troubleshooting steps.

Account anomalies and frequent requests: reduce the chance of triggering risk controls

If you encounter re-verification requirements, interrupted sessions, or “suspicious activity” prompts, ChatGPT troubleshooting recommends stopping high-frequency refreshes and repeated retries, waiting a bit, and then logging in again. Make sure you’re not rapidly switching logins across too many devices at the same time, and try to keep IP changes from happening too frequently.

If you’re sure your account info is correct but you’re still repeatedly blocked, prioritize using the official help channel to submit a description and screenshots; this kind of restriction usually can only be resolved through account-side review, and blindly repeating actions will only prolong recovery time.