How to Fix ChatGPT File Upload Not Working or Upload Limit Reached

If ChatGPT will not accept a file, first determine whether the problem is specific to that file or affects every upload. Try a small supported file in a new chat. If the small file works, check the original file's type, size, and content limits. If every upload fails, check your account limits, OpenAI service status, browser or app, extensions, VPN, and network before deleting data or changing account settings.

An “upload limit reached” message is a different problem from an unsupported file or a temporary network error. OpenAI applies per-file limits, upload-rate limits, and account storage limits. Failed upload attempts can also count toward the rolling upload cap, so repeatedly retrying the same file can make troubleshooting harder.

Identify the Upload Failure Before You Change Anything

The message and timing of the failure usually tell you where to start:

Use a non-sensitive test document rather than repeatedly uploading the original file. That gives you a clean comparison without exposing information unnecessarily.

Check Whether the File Type Is Supported

OpenAI currently supports common document, spreadsheet, presentation, and text formats including PDF, DOCX, PPTX, XLSX, XLS, CSV, TSV, and TXT.

A Google Docs .gdoc shortcut file is not currently supported. Export the Google document as a real file such as PDF or DOCX, then upload the exported copy. Renaming .gdoc to another extension does not convert the document and should not be used as a workaround.

If the extension is supported but the upload still fails, continue to the size and account checks below.

Check the Per-File Limits

OpenAI's current File Uploads FAQ lists several different limits, so a file can fail even when its size looks reasonable:

The 512 MB maximum is not the only limit. A very large text-heavy PDF can hit the document token limit before it reaches 512 MB, while a spreadsheet has a lower practical size limit.

Do not split or compress a file blindly. First keep an untouched original, then make a smaller copy containing only the pages, sheets, or rows needed for the task. Confirm that the smaller copy does not remove information required for the answer.

Run a Small Supported-File Test

Create a new chat and upload a simple TXT or PDF file that contains no private information and is well below the published limits. Then ask ChatGPT to identify one obvious detail from the file.

The result narrows the problem:

This test is more useful than deleting chats or reinstalling software immediately because it separates a file problem from a broader upload problem.

Fix “Upload Limit Reached” Without Deleting Everything

Confirm the Account and Recent Upload Activity

First confirm that you are signed in to the account and subscription you intended to use. OpenAI's current FAQ distinguishes rolling upload-rate limits from storage limits, and it notes that failed upload attempts can count toward the rate cap.

The published rate limit can also vary by plan and system conditions. The current FAQ states that Free users have a lower daily upload allowance and that OpenAI may reduce upload limits during peak periods. Because these limits can change, use the current account interface and official FAQ rather than assuming an old number still applies.

Review Library Storage Separately from Chat History

If Library is available for your account, open Library and use its Storage view to check total usage and remaining storage. Delete only files you no longer need after confirming that you have another copy when necessary.

Chat history and Library files are managed separately when Library is available. OpenAI states that deleting a chat does not delete files saved to Library. If storage is the problem, review the Library itself instead of deleting conversations at random.

Open the Library, locate the file, and delete the specific saved file you no longer need. Files deleted from Library are removed from the main Library view immediately and follow OpenAI's retention process. Do not delete a file that exists nowhere else simply to test whether uploads resume.

Check OpenAI Service Status Before Resetting Your Browser

If multiple files fail across more than one chat, check OpenAI's official status page for an active incident affecting ChatGPT or file uploads. A service-side incident cannot be fixed by clearing local data.

If no relevant incident is listed, refresh ChatGPT or restart the app and try the small test file again. This keeps the troubleshooting order simple and avoids unnecessary destructive changes.

Rule Out Browser, App, and Network Problems

Start with a Clean Browser or App Session

OpenAI's troubleshooting guidance recommends isolating local browser and network causes when errors persist. Use this order:

  1. Start a new chat and retry once.
  2. Refresh the page or restart the ChatGPT app.
  3. Open ChatGPT in an incognito or private window.
  4. Temporarily disable browser extensions that block scripts or modify page content.
  5. Temporarily turn off a VPN, proxy, or secure-DNS filtering tool if one is active.
  6. Try another supported browser or device.
  7. Try a different network.
  8. If the problem continues, clear ChatGPT site data or browser cache and sign in again.

Restore Security Tools After the Comparison

Re-enable security and privacy tools after the comparison test. The goal is to identify the conflicting layer, not to leave protections disabled permanently.

When Only One PDF or Document Keeps Failing

If other files upload normally, preserve the original and create a diagnostic copy. For a PDF, try a smaller page range. For a spreadsheet, copy only the required sheets into a new file. For a presentation, remove unnecessary embedded media from a copy rather than changing the original.

If the smaller copy uploads, the problem is likely related to the original file's size, complexity, or content limit. If both copies fail while unrelated files work, export the document again using the source application's normal export function and retry with the newly created file.

Once uploads work, How to Use ChatGPT Study Mode with Your Notes and PDFs explains a source-first workflow for studying uploaded documents. For table-focused PDF work, How to Extract PDF Tables into CSV and Verify the Data covers verification against the original pages.

Protect Sensitive Files While Troubleshooting

Do not use a confidential contract, customer export, medical record, password list, or other sensitive file as a troubleshooting sample. Create a small dummy file instead.

If you upload real work material, review your ChatGPT Data Controls and your organization's policies first. Personal and managed workspaces can have different settings and retention requirements. A successful upload only confirms that the feature works; it does not mean every file is appropriate to upload.

What Not to Do

Confirm That the Upload Problem Is Resolved

Use the same small test file after each meaningful change. The problem is resolved when the file finishes attaching, appears in the chat, and ChatGPT can answer a simple question based on its contents without an upload error.

Then retry the original file once. If the small file succeeds but the original still fails, stop changing account or browser settings and return to the original file's type, size, export, and content limits.

When to Contact OpenAI Support

If uploads fail across multiple supported files, browsers or apps, devices, and networks while no relevant service incident is listed, collect the information Support can use to investigate:

OpenAI Support may also ask for a HAR file when diagnosing web-interface problems. Follow the official support instructions when collecting it because HAR files can contain sensitive request information.

Official OpenAI References

Use the Failure Pattern to Choose the Next Step

A file rejected immediately needs a file-type or limit check. An “upload limit reached” message needs an account, rate, or storage check. A failure affecting every supported file needs service, browser, app, or network troubleshooting. Keeping those cases separate prevents unnecessary deletion and makes it easier to confirm exactly what fixed the problem.

Related Guides

About the author

Tweaknook Editorial publishes practical guides and browser-based tools for everyday digital work. Product-dependent facts are checked against current primary documentation, with limitations and safer verification steps stated where relevant.