---
title: "Import a data export"
description: "Import the data export from ChatGPT, Claude, Gemini, Microsoft Copilot, Perplexity, Grok, DeepSeek or GitHub Copilot: how to get it, step by step."
url: https://chattomarkdown.com/docs/1.0.0/import/
version: "1.0.0"
updated: 2026-11-04
---

# Import a data export

> Ask the chat service for your data, then drop the file it gives you (the ZIP as it came, a CSV or JSON) on the import page. Every chat in it is saved to your folder at once, without the site's speed limits. On Free, an import fills your 10 free chats with the export's latest chats; Pro imports every chat.

Every chat service lets you download your data. Importing that export saves all the chats in it at once, which is much faster than waiting for a long history to sync. It also works for chats a sync can't reach.

You don't have to import anything: with a service turned on, the extension saves your history from the site by itself. Import when your history is long and a site's limit makes syncing slow, or when you already have an export.

## Open the import page

- In the popup, click **Import past chats** (the download icon at the top).
- Or go to **Settings → Import** and click **Import a data export**.

![Settings, Import section with the Import a data export row](https://chattomarkdown.com/docs/shots/1.0.0/settings-import-light.webp "Settings → Import")

The page is called **Import your chat history**: "Add the data export from any of your chat services. Every chat in it is saved to your folder as Markdown, all at once."

![The import page with the drop area and a tile for each service](https://chattomarkdown.com/docs/shots/1.0.0/import-drop-light.webp "Drop your export, or pick a service for the steps to get it")

You can import before you press **Start saving**. You need a folder; the page asks for one if you haven't chosen it.

## Step 1: get your export

Under **Don't have your export yet?**, click a service for the steps. Each guide has a button that opens the right page on the service's site.

![The import steps for one service](https://chattomarkdown.com/docs/shots/1.0.0/import-guide-light.webp "The steps to get one service's export")

### ChatGPT: a ZIP by email

1. **Open ChatGPT's data controls.** In ChatGPT: **Settings → Data controls**.
2. **Export your data.** Click **Export data**, then **Confirm export**.
3. **Download it from the email.** OpenAI emails you a link, usually within minutes. It works for 24 hours.
4. **Drop the ZIP here.** As it came; no need to unzip it.

### Claude: a file or ZIPs by email

1. **Open Claude's privacy settings.** In Claude: **Settings → Privacy**.
2. **Export your data.** Click **Export data**.
3. **Wait for Claude's email.** It arrives within minutes to hours, with a link to a small file, or to the ZIPs. Its links work for 24 hours.
4. **Drop the file here.** Drop the file from the email, or the ZIPs you downloaded.

If you drop the small file, the page says **Download your Claude export** and asks you to allow claude.ai. Click **Allow claude.ai and download**: the extension downloads your chats from Claude, signed in as you are in this browser. Claude's export names the images and files in a chat but doesn't hold them; with claude.ai allowed, they come from there.

> [!NOTE]
> The note under that button says allowing claude.ai also turns Claude on. In this version it doesn't: it only allows the download. To have new Claude chats saved on their own, turn Claude on in **Settings → Chat services**.

### Gemini: Google Takeout

Google emails the Takeout within hours to a day.

1. **Open Takeout with only My Activity ticked.** Click **Open Google Takeout**. If everything is ticked, click **Deselect all** and tick **My Activity**. Not the item called **Gemini**: it holds only Gems and scheduled actions.
2. **Keep only Gemini Apps.** Under My Activity, click **All activity data included**, then **Deselect all**, tick **Gemini Apps** and click **OK**.
3. **Choose JSON.** Click **Multiple formats** and set Activity records to **JSON**. HTML can't be read.
4. **Create the export.** Click **Next step** and **Create export**, then download the ZIP from Google's email.
5. **Drop the ZIP here.** As it came.

No Gemini Apps under My Activity? Check that Takeout is open in the Google account you use Gemini with, and that Gemini Apps Activity is on.

### Microsoft Copilot: CSV files, right away

1. **Open Microsoft's privacy dashboard.** Its **Copilot** page, signed in with the Microsoft account you use Copilot with.
2. **Export every Copilot section.** Click **Export all activity history** under each Copilot section the page lists. Each downloads a CSV.
3. **Drop the CSVs here.** All of them together. A file with only its column names means Microsoft has no chats there.

Microsoft's export can come back empty even when Copilot shows your chats, and it never has chats from while you weren't signed in. Saving straight from Copilot reads every chat in its sidebar.

### Perplexity: a ZIP

1. **Open your account settings.** In Perplexity: **Settings → Account**.
2. **Export your data.** Under **Data export**, click **Export my data**.
3. **Drop the ZIP here.** Download it when it's ready, and drop it as it came.

Perplexity's export has only the text of each thread. Allow perplexity.ai when you import, and the sources and pictures come from there.

### Grok: a ZIP by email

1. **Open your account data.** In Grok: **Settings → Data Controls**.
2. **Download your data.** Click **Download account data**.
3. **Drop the ZIP here.** Download it from the email when it comes, and drop it as it came.

### DeepSeek: a ZIP

1. **Open DeepSeek's settings.** In DeepSeek, click your profile at the bottom left, then **Settings**.
2. **Export your chats.** Under **Data management**, click **Export** next to Export all conversations.
3. **Drop the ZIP here.** Download it when it's ready, and drop it as it came.

### GitHub Copilot: a JSON file per chat

1. **Open a chat on GitHub Copilot.** At github.com/copilot.
2. **Export it.** Open the chat's menu, then **Export → JSON**.
3. **Drop the files here.** One per chat. Drop as many as you like, all at once.

### Or save them straight from the site

Each guide also has a card: "No export or waiting: every chat on chatgpt.com, with its pictures and files, kept up to date from then on." Its button turns the service on (**Turn on ChatGPT**), or, when it is on, saves its chats now (**Save them now**). On Free that is up to 10 chats in all.

## Step 2: drop the file

Drop the export anywhere on the page, or click **Choose file**. The page reads:

| From | Drop |
|---|---|
| ChatGPT | The ZIP |
| Claude | The small file from the email, or the ZIPs (several at once) |
| Gemini | The Takeout ZIP (several parts at once) |
| Microsoft Copilot | The CSV files, together |
| Perplexity | The ZIP, or its Markdown threads |
| Grok | The ZIP |
| DeepSeek | The ZIP |
| GitHub Copilot | The JSON files |

The page finds out which service an export is from by what is inside, not by its name. Drop one export at a time. ZIPs of any size work; a single CSV, JSON or Markdown file can be up to 256 MB.

While it reads, the page says "Reading" with the file name and how much it has read. Your export is read on your computer and never uploaded.

## Step 3: check what was found

![The import page after reading a ChatGPT export on Pro: the chats found and the folder they go to](https://chattomarkdown.com/docs/shots/1.0.0/import-found-light.webp "On Pro: every chat in the export")

The page says how many chats it found, for example "1,204 chats found in your ChatGPT export", and shows:

- **Where they go**: your folder and the service's folder inside it, "Ready to save here". If you haven't chosen a folder, click **Choose folder**. If Chrome needs your OK again, click **Allow access**.
- **What comes from where**: some exports leave things out, such as uploads (ChatGPT, Claude) or sources and pictures (Perplexity, Gemini). A card lists them as "From the export", "From chatgpt.com" or "Needs access". Click **Allow access** to let the extension fetch the rest from the site, with your sign-in in this browser. Without it, the button says ", text only" or ", export only", and only what the export holds is saved.
Click **Import 1,204 chats** to start.

### On the Free plan

![The import page on Free: the latest chats that fit in your 10, or every chat with Pro](https://chattomarkdown.com/docs/shots/1.0.0/import-found-free-light.webp "On Free, with 7 of 10 free chats used")

On Free the page offers a choice, with a chart of your history by month: **Your 10 latest chats** (or as many as you have left, saved now) or every chat in the export with Pro. Click **Import free** to save the ones that fit. See [Free and Pro](https://chattomarkdown.com/docs/1.0.0/free-and-pro/).

## Step 4: wait for it to finish

![The import in progress: chats saved, then images and files](https://chattomarkdown.com/docs/shots/1.0.0/import-progress-light.webp "Chats first, then their images and files")

The import has two steps:

1. **Chats**: every chat is written to your folder, with a count and the time left.
2. **Images and files**: the pictures and files of those chats, from the export or from the site.

The page counts **New**, **Updated** and **Already saved** chats as it goes. Keep the tab open until it is done. Once the chats are saved, you can click **Skip files**; importing the same file again later adds the rest. **Cancel** stops after the chat being saved.

If a sync is writing to your folder at that moment, the page says "Waiting for a sync to finish with the folder. The import goes on right after."

## Step 5: done

![The import finished, with counts and the newest imported chats](https://chattomarkdown.com/docs/shots/1.0.0/import-done-light.webp "Done: what was saved")

The page says, for example, "1,204 chats imported from ChatGPT", with the counts, how many images and files were added, and the newest chats from the import. **Open in the reader** shows them; **Import another file** starts again.

If some images or files couldn't be added, a list says why: "Bigger than 25 MB", "Not in the export or on chatgpt.com", or "Not in the export. Allow chatgpt.com to download it".

Your next sync picks up anything newer than the export.

## What happens to chats you already saved

- A chat you saved before is **updated in place**, never doubled.
- If your saved copy is newer than the export's (the sync saved it after the export was made), it is **left as it is**.
- If it is the same, it counts as **Already saved**. Missing pictures are still added.
- Notes you wrote above or below a chat stay, as with any update.
- For Gemini, Microsoft Copilot and Perplexity, whose exports hold less than the site, a chat the sync already saved keeps its file.
- Your skip rules apply, as in a sync. A file whose markers you removed is skipped.

## Free and the import

On Free, an import saves the export's latest chats until your 10 free chats are used, from all services together. Chats that are already among your 10 are kept up to date and don't count again. A chat counts as soon as the import reaches it, even if you stop the import.

When the export has more chats than you have left, the done view says how many are still in your export and offers Pro for the rest. If you upgrade while the page is open, it says "Pro is on. Import the other 1,194 chats" with **Import 1,194 more**. See [Free and Pro](https://chattomarkdown.com/docs/1.0.0/free-and-pro/).

## If the import stops

If you close the tab, Chrome closes, or you click **Cancel**, nothing saved is lost. The import page then shows **Continue your ChatGPT import** with how far it got.

1. **Open the import page again.**
2. **Click Continue import.** If Chrome can't reopen the file by itself, the page asks you to choose the same file again.
3. **Let it finish.** It picks up where it stopped; nothing is saved twice.

**Discard** forgets the stopped import. Starting a different import also replaces it.

## Problems

| Message | What to do |
|---|---|
| "These chats don't look like an export Chat to Markdown can read" | Drop the file the service gave you, as it came: the ZIP, the CSVs, or the JSON files. |
| "Drop one export at a time." | Import each service's export on its own. Only parts of one export (Takeout, Claude) go together. |
| "This Takeout has your Gemini chats as HTML, which can't be read." | Export again with Activity records set to **JSON**. |
| "This Takeout has Gemini's Gems and settings, not your chats" | You picked the item called Gemini. Export **My Activity → Gemini Apps** instead. |
| "This Copilot export is empty" | Check you are signed in with the Microsoft account you use Copilot with, and export every Copilot section. |
| "The links in this file have expired: they last 24 hours." | Request a new export in Claude's settings, or drop the ZIPs you downloaded. |
| "You're signed out of claude.ai." | Sign in to Claude in this browser, then try again. |
| "The ZIP file is incomplete. Download the export again." | The download was cut short. Download it again. |
| "Not signed in to chatgpt.com, so only what the export holds is added." | Sign in to the site in this browser and import the same file again for the rest. |
| "All 10 free chats are used. Pro saves every chat." | Free is full. Your 10 stay up to date. |

More fixes: [Troubleshooting](https://chattomarkdown.com/docs/1.0.0/troubleshooting/).

## Questions

### Do I have to unzip the export?

No. Drop the ZIP as it came. An export in several parts (Google Takeout, Claude) can go in together.

### Is my export uploaded anywhere?

No. It is read on your computer and the chats go only to your folder.

### Will importing duplicate chats I already saved?

No. Chats you saved before are updated in place. A chat the sync saved after the export was made is left as it is.

### Can I import on Free?

Yes. On Free, an import saves the export's latest chats until your 10 free chats are used. Pro imports every chat in it.
