What is the official ChatGPT data export?
ChatGPT provides an account data export flow that lets you request a copy of your data. The downloaded archive can contain a file named conversations.json, which stores conversation records in a structured format. This page reads that file in your browser so you can select chats and prepare a separate archive.
This workflow is designed for people who want to save more than one conversation. It is different from exporting a single public share link.
How to request your ChatGPT data
- Open ChatGPT and go to Settings.
- Open Data Controls and look for Export Data.
- Confirm the export request.
- Follow the instructions in the email from OpenAI to download the archive.
- Unzip the downloaded archive on your device.
- Find
conversations.jsonand choose it in the tool above.
Menu names and availability can change. Follow the current instructions shown in your ChatGPT account.
Select conversations and create a ZIP archive
After you choose the JSON file, the tool lists the conversations it can read. Select the conversations you want to keep and choose an output format. Download selected creates one ZIP file containing a separate file for each selected conversation. This keeps individual chats separate instead of merging unrelated conversations into one document.
The browser reads the file locally. The selected data is not sent to the ChatExport server by this page. Make sure you are using a trusted device if the archive contains sensitive or work-related conversations.
What can affect the result?
- Conversation branches: the parser follows the active conversation path represented by the current node. Alternate branches may not be included.
- Non-text messages: messages that contain only images or other non-text content may not appear in text exports.
- Large archives: processing a very large JSON file can use significant browser memory.
- Format differences: JSON exports can change over time. If a file is not recognised, download a fresh archive and check that you selected
conversations.json.
People Also Ask: Common Questions and Detailed Answers
What does this ChatGPT tool actually export?
This page is designed for people who want to keep a usable copy of material that appears in a ChatGPT conversation. It works with the conversations.json file from your official ChatGPT data archive. The tool focuses on readable message content and keeps the conversation title and speaker order when those details are available. Depending on the selected workflow, you can create TXT, Markdown, JSON, HTML or PDF files inside a ZIP archive. It is a practical export or cleanup utility, not a copy of the complete ChatGPT application and not a way to retrieve content that you are not permitted to access.
Some source elements may not be represented in a text-based export. Examples include attached files, generated images, audio, video, interactive controls, hidden metadata and alternate response branches. For an account-level backup, keep the original official data archive. For an everyday document, choose your chosen file format and inspect the downloaded result before relying on it. This distinction helps you choose the right file for the job instead of expecting one format to preserve every feature.
How do I use this tool from start to finish?
Begin with the conversations.json file from your official ChatGPT data archive. If you use a public share link, copy the complete URL from ChatGPT and make sure it opens in a logged-out browser window. If you use pasted text, include speaker labels such as You: and ChatGPT: when they are available. If you use an account archive, extract the ZIP on your own device and select the correct JSON file. Load the source, wait for the preview or parser to finish, and check the conversation title and message count before exporting.
Next, select the messages or conversations you want to keep. Choose your chosen file format, then start the export. Wait for the browser download to finish and open the file. Check the first and last messages, speaker order, long paragraphs, code blocks, lists, tables and special characters. If the output is intended for another person, remove unrelated material and choose a clear filename. A final review catches missing content before the file becomes part of a report, project folder or shared archive.
Can I export a private ChatGPT conversation using a public link?
A public share link only represents a conversation page that has been made accessible through ChatGPT sharing. It does not grant access to your private account history, and this tool does not sign in to your ChatGPT account. If a conversation is private, request the official account data export from ChatGPT settings. Download the archive, extract it on your device and use the relevant local import workflow if the tool provides one. If you only need a short excerpt, you can copy the text you are allowed to access and use the paste fallback.
Do not make a sensitive conversation public solely to use an exporter. A share link can be forwarded or opened by people who receive it, depending on the sharing settings. Check the access level and review the content before sharing. If you are exporting work, customer or research material, follow the rules of the organisation or project that owns the information. The safest workflow is the one that matches the original privacy level and the permissions you have.
Will the export keep code blocks, tables, lists and formatting?
Formatting depends on the source and destination. message roles, order, structured fields and machine-readable data are the main things to check for this page. When the source page exposes structured message content, the extractor converts rich elements into text and Markdown where possible. When you paste text manually, the tool can only work with the formatting that was included in the clipboard content. An export cannot reconstruct a table or code sample that was already lost before the text reached the tool.
Open the downloaded file in the application where you plan to use it. Markdown viewers do not all support identical extensions, spreadsheet-like tables may need adjustment, and long code lines can wrap differently in a PDF. For complex scripts such as Hindi or Arabic, browser print-to-PDF can use fonts installed on your device. Keep the original archive when exact structure or non-text attachments matter. Consider exporting both an editable file and a reader-friendly version for important conversations.
What should I do if a share link or file does not load?
For a share link, open it in a private browser window and confirm that the page is public, complete and still available. Check for a missing character in the URL and make sure it belongs to an accepted ChatGPT share domain. Automated access may occasionally fail when the source platform changes its page structure or limits requests. If the page opens manually but extraction fails, try the paste fallback for text you can view, or use the official account archive for a private or larger backup.
For an uploaded JSON file, verify that you selected the extracted conversations.json file rather than the ZIP archive or another JSON document. A very large file can take time and memory to parse. Close unnecessary tabs and retry with a smaller selection if the browser becomes slow. If the problem continues, keep the original source untouched and use a local JSON editor or trusted script to validate the file. Do not repeatedly submit a link that the source website is blocking.
Can I choose individual messages instead of exporting everything?
When a conversation preview is available, use the message checkboxes to include only the exchanges you need. This is useful when one long chat contains several unrelated subjects or when you need to share a focused excerpt with a colleague. Keep the question that prompted an answer and any corrections or caveats that change its meaning. A response without its surrounding context can be misunderstood, so select messages as a complete sequence rather than simply choosing the shortest output.
For bulk imports, select only the conversations relevant to the task before building the ZIP. If the file will be used as a formal record, state whether it is a complete transcript or a selected extract. Read through the selection once more before downloading, especially when references such as "the earlier example" or "option two" depend on a message that may have been left out. Selective export helps reduce clutter, but it should not remove information required to understand the result.
Does ChatExport store my conversation or uploaded archive?
The public-link extractor needs to retrieve the shared page on the server before it can return the visible message content to your browser. The service does not provide an account-based conversation library or a saved archive feature. The bulk data import and paste-based cleanup workflows are designed to process selected text or JSON locally in the browser. Technical infrastructure may still process request metadata such as IP address, timestamps and error logs to operate and protect the service. Read the Privacy Policy for the current processing description.
Do not assume that a web utility is suitable for every confidential conversation. Avoid submitting information that you are not authorised to share, and remove personal or sensitive material before creating a public link or sending a downloaded file. If your organisation has rules about AI data, follow those rules. For sensitive records, use an approved local workflow and preserve the original file in storage controlled by you or your organisation.
Which format should I choose and what should I check afterward?
Choose a format based on the next action. PDF is useful for scripts, data review, repeatable backups and structured imports only when a fixed page layout is important. TXT is convenient for quick search and simple backups. Markdown is a good choice for editable notes and technical documentation. JSON is useful when a script or another application needs structured message fields. HTML is convenient when you want a local page that opens in a browser. The best option depends on whether readability, editability, portability or machine processing matters most.
After exporting, open the file rather than relying only on the success message. Confirm the title, message count, speaker order, line breaks and special characters. For PDFs, inspect page breaks and the final page. For Markdown, check code fences and tables. For JSON, validate the file with a JSON-aware editor. For HTML, open it offline and check links. Keep the original archive or source copy for important material, because a converted export is a simplified representation of the conversation.