ChatGPT file upload failures have three main sources: plan restrictions (file uploads require ChatGPT Plus or higher), file-level issues (wrong type, too large, corrupt), and browser or network problems (upload gets blocked before reaching OpenAI). Identifying which applies takes under two minutes. Work through these checks in order — most issues resolve without contacting support.
Paperclip or attachment icon is grayed out or missing from the ChatGPT message bar
File appears to upload (progress bar shows) but then disappears with no confirmation
Error message: 'Failed to upload file. Please try again.'
Error message: 'Unable to read file' or 'Could not process this document'
ChatGPT accepts the file but then says 'I cannot see the contents of this file'
Upload progress bar stalls at a specific percentage and never completes
PDF uploads work but ChatGPT only reads part of the document and misses later pages
File upload works on desktop but not in the ChatGPT iOS or Android app
File uploads (PDFs, images, data files) in ChatGPT require ChatGPT Plus ($20/month) or higher. Free tier users see the attachment icon either grayed out or absent entirely. If you're on free tier, upgrading to Plus is the only path to file uploads — there is no workaround within ChatGPT itself.
ChatGPT supports: PDF, DOCX, PPTX, TXT, CSV, JSON, XML, HTML, most code file types (.py, .js, .ts, .java, .cpp etc.), and common image formats (JPG, PNG, GIF, WEBP). It does NOT support: XLSX directly (convert to CSV), video files, audio files, ZIP/archive files, or executable files. Attempting to upload unsupported types causes silent failures or generic upload errors.
ChatGPT enforces a 512 MB per file maximum for most file types, but text-based processing (reading content) is limited by the model's context window. PDFs with more than ~300 pages of dense text may upload successfully but produce 'could not process' errors during analysis. Images are limited to 20 MB per image. Files approaching size limits are more likely to fail during upload on slower connections.
Password-protected PDFs, PDFs with DRM (Digital Rights Management), and PDFs that are essentially scanned images (not text-based) all fail ChatGPT's document parser. A PDF that looks readable in Acrobat may fail in ChatGPT if it uses non-standard encoding. Scanned PDFs require OCR processing — ChatGPT can see scanned pages as images but extracts text less reliably than from native PDF text.
Content security policies on some corporate browsers, aggressive ad blockers, or privacy extensions can intercept file upload requests before they leave your browser. This causes uploads to silently fail or produce network errors. This is especially common on Chrome with multiple extensions installed or on company-managed browsers with restrictive security policies.
File uploads are more sensitive to network quality than text messages. A connection that feels 'fine' for browsing can cause upload timeouts for multi-MB files. This is especially noticeable on mobile data, congested WiFi, or corporate VPNs that throttle large uploads. Files that fail on one network often succeed immediately when switching to another.
When to try: First — confirm this before any other troubleshooting
Log into chat.openai.com and look at the message input bar. If the attachment (paperclip) icon is missing or grayed out, your current plan doesn't include file uploads. Free tier users must upgrade to ChatGPT Plus to access file upload. To upgrade: click your profile icon (bottom left) → 'Upgrade plan' → select Plus ($20/month). File uploads, image analysis, and code interpreter all unlock immediately after subscribing.
When to try: When uploads fail silently or with 'unsupported file type' error
Verify your file is a supported type: PDF, DOCX, PPTX, TXT, CSV, JSON, HTML, most image formats (JPG, PNG, WEBP), or code files (.py, .js, .ts, etc.). If you have an Excel file (.xlsx): open it → File → Save As → CSV. If you have an audio or video file: ChatGPT cannot read these natively — transcribe audio with Whisper first. If you have a ZIP archive: extract it and upload files individually.
When to try: When upload stalls or fails with very large files
For PDFs: if the file is over 25 MB, compress it first. On Mac: open in Preview → File → Export as PDF → set Quartz Filter to 'Reduce File Size'. On Windows: Adobe Acrobat → File → Save as → Reduced Size PDF. For images: resize to under 10 MB using any image editor. If your PDF has hundreds of pages, split it into sections: on Mac use Preview → select pages → drag to Finder. Adobe Acrobat (free online at acrobat.adobe.com) also splits PDFs.
When to try: When ChatGPT says 'I cannot read this file' or 'unable to process document'
ChatGPT cannot process password-protected or DRM-locked PDFs. To remove password protection: open the PDF in Chrome browser → enter the password when prompted → File → Print → change printer to 'Save as PDF' → save. The new file will be unlocked. For scanned PDFs (image-only, no selectable text): use Adobe Acrobat's OCR feature (Tools → Enhance Scans → Recognize Text) to create a text-searchable version before uploading.
When to try: When upload fails in normal browser but you suspect an extension conflict
Open a new incognito/private window (Ctrl+Shift+N in Chrome, Cmd+Shift+N on Mac), log into chat.openai.com, and attempt the upload again. Incognito disables all extensions. If the upload works in incognito, an extension is blocking it in your normal browser. Likely culprits: uBlock Origin, Privacy Badger, or corporate security extensions. Disable extensions one at a time in normal mode to find the conflict.
When to try: When upload stalls mid-way or times out on specific network
Turn off any active VPN before uploading. If on corporate WiFi, try from your phone's mobile hotspot. File uploads route through different OpenAI endpoints than chat messages, and some networks block them specifically. Test: if a 1 MB text file uploads fine but a 10 MB PDF doesn't, the issue is likely network throttling on large transfers. Once the file is uploaded, you can re-enable your VPN and continue chatting.
When to try: Specifically for ChatGPT mobile app upload failures
On iOS: tap the + button to the left of the message field → choose 'Files', 'Photos', or 'Camera'. If the button is missing, update the ChatGPT app: App Store → your profile → look for ChatGPT under pending updates. On Android: tap the + or paperclip icon → select file source. If app crashes during upload: clear the app cache (Settings → Apps → ChatGPT → Storage → Clear Cache), then relaunch. For large files on mobile, WiFi is strongly recommended — mobile data uploads frequently stall on files over 5 MB.
When to try: When uploads fail in one browser but you suspect a browser-specific issue
If Chrome is failing, try Microsoft Edge or Firefox. ChatGPT file uploads work across all modern browsers, but browser-specific issues (corrupted browser profiles, broken IndexedDB storage used for upload buffering) can cause persistent failures. In Chrome: try creating a new browser profile (Profile icon → Add profile → Continue without signing in) and test uploads there. A fresh profile isolates whether the issue is Chrome-wide or specific to your profile.
Convert Excel files to CSV before uploading — ChatGPT handles CSV far more reliably than XLSX
Compress large PDFs before uploading to avoid timeout failures on slower connections
Remove password protection from PDFs before uploading — do this once and save the unlocked version
For scanned documents, run OCR before uploading to get accurate text extraction
Keep the ChatGPT mobile app updated — upload-related bugs are frequently fixed in app updates
Split large documents (100+ pages) into logical sections for more reliable uploads and better analysis quality
Contact OpenAI support at help.openai.com if: (1) You're on ChatGPT Plus and the attachment icon is still missing or grayed out after signing out and back in, (2) You receive a specific error code in the upload failure message (share that code with support), (3) File uploads worked previously and have stopped for all file types despite trying multiple browsers and networks, (4) You're on ChatGPT Team or Enterprise and uploads are broken for your whole team. For mobile app crashes during upload that persist after reinstalling, report via the in-app feedback button (Profile → Share Feedback) with the specific file type and size that fails.
No. File uploads — including PDFs, images, data files, and code files — require ChatGPT Plus ($20/month) or a higher plan. Free tier users do not see the attachment icon, or see it grayed out. The only way to share document content on free tier is to copy-paste the text directly into the chat (limited by the context window, roughly 300 pages for GPT-4o). Upgrading to Plus enables file uploads immediately.
Supported types as of April 2026: Documents (PDF, DOCX, PPTX, TXT, RTF), Data files (CSV, JSON, XML), Code files (.py, .js, .ts, .java, .cpp, .c, .rb, .go, .html, .css, and most other source code extensions), Images (JPG, PNG, GIF, WEBP). Not supported: Excel XLSX (convert to CSV first), audio files, video files, ZIP archives, and executable files (.exe, .dmg). When in doubt, save as PDF or CSV — both are reliably processed.
Four common reasons: (1) The PDF is password-protected or DRM-locked — remove the password first (open in Chrome, print to PDF). (2) It's a scanned PDF (image-only, no selectable text) — ChatGPT can see it as images but text extraction is less reliable. (3) The PDF uses non-standard encoding — try re-saving it from Acrobat or Chrome's print-to-PDF. (4) The document is too long for the context window — very long PDFs may upload but ChatGPT can only analyze the portion that fits in 128K tokens.
The technical limit is 512 MB per file, but practical limits are lower. For text-based processing, you're limited by the context window — ChatGPT can read roughly 300 dense pages before hitting the 128K token limit of GPT-4o. Images should be under 20 MB for reliable upload. Large files are more likely to time out on slow connections — if a file is over 25 MB, compress it first. For data analysis files (CSV, JSON), under 50 MB is the practical sweet spot.
The file uploaded successfully to OpenAI's servers, but ChatGPT's analysis step failed separately. Most common causes: the file is an image-scanned PDF with poor OCR quality, the document is in a language with unusual encoding, or the file is too long for the context window and ChatGPT only processed part of it. Fix: try attaching the file again in a new message with a specific question ('Summarize the key points from pages 1–20 of this document') rather than just uploading and waiting.
Yes, ChatGPT Plus supports uploading multiple files per message. Click the attachment icon and select multiple files, or drag and drop them together onto the message area. There's no published limit on the number of files per message, but total context (all files + conversation history) must fit within the 128K token window. For large-scale document analysis across many files, Claude Projects or NotebookLM are better suited — they handle multi-document reference more cleanly.
Mobile uploads have additional failure modes: (1) Mobile data connection drops during upload for files over 5 MB — switch to WiFi. (2) The ChatGPT app is outdated — update it via the App Store or Play Store. (3) iOS file picker or Android file manager doesn't have permission to access the folder where your file is stored — check Files app permissions. (4) Some file types that work on desktop (DOCX) occasionally fail on mobile due to different MIME type handling — try uploading as PDF instead.
The upload itself doesn't consume a message. The message that includes the attached file (where you ask ChatGPT to analyze it) counts as one message toward your hourly quota. Processing large documents is compute-intensive, so some Complex file-analysis messages may count more heavily against limits than simple text messages. If you're near your limit, finish your file analysis before switching to other tasks. Verified April 2026.