Recovering Your Data from ChatGPT Business Accounts for Free

If you are transitioning away from a ChatGPT Business or Team workspace, you may have noticed a frustrating omission: ~~the "Export data" button~~. While personal accounts enjoy a simple one-click archive, business tiers lack an official mechanism to retrieve conversation histories before a subscription ends or a workspace is deactivated.

scrapemychats is an open-source solution designed to bridge this gap. It allows you to migrate every conversation—including all text and attached files—directly to your local hardware, generating a sophisticated, searchable, offline archive.

Note: While designed for Business/Team accounts, this tool is fully compatible with personal accounts as well.

🛡️ Privacy and Security

Your data remains strictly local. The tool operates by controlling a legitimate Google Chrome window on your machine using your existing active session.

No Password Access: The script never interacts with or sees your password.
No External Servers: Traffic is routed exclusively to chatgpt.com.
Local Storage: Your history never leaves your physical machine.

📂 The Export Architecture

The tool organizes your data into a structured directory. Here is the breakdown of the export/ folder:

File/Folder	Purpose	Contents
`viewer.html`	Main Interface	The primary entry point for searching and browsing.
`manifest.csv`	Index	A spreadsheet containing status and counts for every chat.
`errors.log`	Debug Log	A record of any chats that failed to download.
`001_ChatTitle_ID/`	Chat Folder	Individual directories for every single conversation.

Inside each individual chat folder, you will find:

conversation.json: The raw data (including tool calls and metadata).
conversation.md: A human-readable Markdown transcript.
files/: A subfolder containing all images and generated attachments.

🖥️ The Offline Viewer

The generated viewer.html is a standalone file that requires no server or internet connection—making it ideal for storage on a USB drive.

Instant Search: Full-text search that highlights matches in real-time.
Organization: A three-pane layout (Categories $\rightarrow$ Chat List $\rightarrow$ Conversation).
Rich Media: Inline thumbnails for images and full table support.
Navigation: Full keyboard support for rapid browsing.

⚙️ Technical Workflow

The process follows a linear logic path: $\text{Authentication} \rightarrow \text{Discovery} \rightarrow \text{Extraction} \rightarrow \text{Indexing} \rightarrow \text{Visualization}$

🚀 Quick Start Guide

Prerequisites: Python 3.10+ and Google Chrome (or MS Edge via --browser-channel msedge).

Test Run: Verify the setup with a small sample.
```
python export_chats.py --limit 3
```
Full Export: Download the entire history.
```
python export_chats.py
```
Generate Viewer: Create the HTML interface.
```
python build_viewer.py
```

Tip: The tool automatically finds chats (including those in "Projects"), but you can provide a custom list using --csv mylist.csv.

🛠️ Detailed Installation Instructions

For Windows Users

Install Python: Download from python.org. Crucial: Check the box $\text{Add python.exe to PATH}$ during installation.
Install Chrome: Ensure Google Chrome is installed.
Get the Code: Click the green Code button on GitHub $\rightarrow$ Download ZIP $\rightarrow$ Extract All.
Open Terminal: In the extracted folder's address bar, type powershell and hit Enter.
Install Dependencies:
```
pip install -r requirements.txt
```
Execute Test: Run python export_chats.py --limit 3, log in to the correct workspace, and verify the export/ folder.
Full Run: Run python export_chats.py.

Warning: Large accounts may take several hours. If the process is interrupted or the PC sleeps, simply restart the command to resume from where it left off.
Finalize: Run python build_viewer.py and open export\viewer.html.

For macOS Users

Install Python: Use the macOS installer from python.org.
Install Chrome: Ensure Google Chrome is installed.
Get the Code: Download the ZIP from GitHub and unzip it in your Downloads folder.
Open Terminal:
```
cd ~/Downloads/scrapemychats-main
```

Install Dependencies:

python3 -m pip install -r requirements.txt

Execute Test: Run python3 export_chats.py --limit 3 and log in.
Full Run: Run python3 export_chats.py (leave running in the background).
Finalize: Run python3 build_viewer.py.