ChatGPT FAQ: Messages Won’t Send, Image Uploads, and Rate Limiting

When using ChatGPT, the most infuriating thing is often not “not knowing how to use it,” but suddenly being unable to send messages, failing to upload images, or being told you’re making requests too frequently. Below is an FAQ that consolidates the most common issues and provides a practical troubleshooting sequence by symptom. You can start checking from the easiest step.

Message send failure: endless loading spinner or an error prompt

If ChatGPT won’t send your message, first refresh the page or restart the app, then resend the same content; many temporary congestion issues can be bypassed right away. If it still fails, it’s recommended to copy the prompt text locally, log out and log back in to avoid a stuck session state.

On the browser side, you can also clear the site’s cache and cookies, and temporarily disable extensions such as ad blockers or script managers; these extensions sometimes block ChatGPT requests. For the network, switch to a more stable connection if possible, and avoid frequently switching networks, which can interrupt requests.

Image upload failure: stuck uploading or “not supported”

If ChatGPT fails to upload images, first check the format and file size: a common approach is to convert the image to JPG/PNG and compress it appropriately, then upload again. If you’re uploading multiple images at once, try validating with a single image first; once it works, gradually increase the number.

If it’s still stuck, switching the upload entry point can also help—for example, use drag-and-drop instead of the upload button in the chat box, or switch to another browser/incognito window. On mobile, make sure the app has permission to access your photo library, and turn off Data Saver/Low Data Mode to avoid the system restricting uploads.

Too many requests: rate-limited or asked to try again later

When ChatGPT says you’re making requests too frequently, it’s usually because you’ve sent messages repeatedly in a short time, retried over and over, or asked for a very long output in one go. The most reliable approach is to pause for a few minutes and continue, while splitting the question into two or three parts to reduce the load per generation.

In practice, you can ask ChatGPT for an outline first, then have it expand section by section; this both reduces the chance of triggering rate limits and makes it easier for you to verify the direction. Don’t hammer “Regenerate” repeatedly—this usually just makes rate limiting kick in sooner.

Abnormal answers: repetition, going off-topic, or lost context

If ChatGPT starts repeating itself or drifting off-topic, adding a line in the same turn like “Answer only according to the following three points / don’t restate the background” is often faster than starting a new chat. If the context seems lost, repost the key constraints (goal, format, confirmed information) as 3–5 bullet points to realign ChatGPT.

Also, the longer a conversation gets, the more likely it is to become “fuzzy” later on. Periodically ask, “Please summarize the current conclusions and list the next to-dos,” then continue the discussion based on that summary for better stability. This saves time and makes ChatGPT’s output more controllable.