If you’ve used ChatGPT for a while, it has learned a lot about you. It knows your name, your job, your projects and how you like your answers written.
That is why switching to Claude can feel scary. Nobody wants to start over and explain everything again to a brand-new assistant.
Here’s the good news. You don’t have to. Claude has a built-in way to bring in what ChatGPT knows about you. You copy a list out of ChatGPT, paste it into Claude, and Claude picks up where ChatGPT left off.
My name is Damon Nelson. I build marketing software for small business owners, and Claude is the AI I use every day. Below is exactly how to make the move, one step at a time.
This guide is based on the ChatGPT to Claude guide from Orgrowth and the instructions in Anthropic’s help center. Anthropic is the company that makes Claude.
First, what is “memory”?
Memory is a set of notes an AI assistant keeps about you. Every time you tell ChatGPT something like “I run a bakery” or “always keep your answers short,” it can save that as a note. Then it uses those notes in future chats, so you don’t have to repeat yourself.
Claude has memory too. The steps below move your notes from ChatGPT into Claude.
Step 1: Open Claude on a computer
Do this on a laptop or desktop, not your phone. According to Anthropic’s help center, memory import works on the Claude website and in the Claude Desktop app. It is available on the Free, Pro, Max and Team plans.
Don’t have a Claude account yet? Go to claude.ai and sign up for the free plan. Memory import is included.
Step 2: Ask ChatGPT to list everything it remembers
Open a new chat in ChatGPT. Copy the message below and paste it in.
I’m switching to a different AI assistant and want to take my information with me. Please list every memory you have saved about me, plus anything you’ve learned about me from our past conversations. Put everything in one code block so I can copy it easily. Start each entry with the date it was saved, if you have it.
Be sure to include:
1. Any instructions I’ve given you about how to answer me, like tone, format, style, or things to always or never do.
2. Personal details, like my name, where I live, my job, my family and my interests.
3. My projects, goals and topics I bring up often.
4. The tools, apps and programming languages I use.
5. Preferences and corrections I’ve given you.
6. Anything else you have saved that doesn’t fit above.
Use my exact words where you can. Don’t summarize, group or leave anything out. After the code block, tell me whether that is everything or whether there is more.
ChatGPT will answer with a long list inside a gray box. That gray box is called a code block, and it has a Copy button in the corner.
If ChatGPT says there is more, type Keep going and send it. Repeat until ChatGPT says the list is complete.
Step 3: Read the list and remove anything private
Before you move the list anywhere, read it. You might be surprised by what ChatGPT has saved.
Anthropic suggests leaving out sensitive information. So if you see something you would rather Claude not keep, like health details, money details or private family information, delete that line. The easy way is to paste the list into a notes app, delete the lines you don’t want, and copy the cleaned-up list from there.
Step 4: Save a backup copy
Before you leave ChatGPT, keep a copy of the list for yourself. Anthropic calls memory import “experimental,” which means some entries may not make it across. Your backup makes sure nothing is lost for good.
In the same ChatGPT chat, send this:
Now give me that exact same list as a Markdown file I can download, named my-ai-memory.md.
A Markdown file is just a simple text file. If ChatGPT gives you a download link, save the file. If it doesn’t, paste the list into a note on your computer and save that instead.
Step 5: Find the import button in Claude
Claude has two versions of its settings screen right now, and the button is in a different spot on each one. Check whichever one you see.
- Newer version: click Settings, then Memory, then Start import.
- Older version: click Settings, then Capabilities. Find the Memory section and click Start import.
- Or look on the home screen: you may see a card that says Import memory to Claude with a Get started button.
All three lead to the same screen. It shows a box where you paste your list.
Step 6: Paste your list and click “Add to memory”
Paste your cleaned-up list into Claude’s import box. Then click Add to memory.
Claude reads the list, pulls out the important details and saves each one as its own memory. You can look through them later.
Step 7: Check what Claude learned
On the newer settings screen, Claude’s memory updates shortly after you import. On the older one, it can take up to 24 hours.
When it’s ready, click See what Claude learned about you. That opens a chat with this message already filled in:
I updated my memory. What did you learn about me?
Read Claude’s answer carefully. If anything is wrong, fix it now so Claude starts off with the right information.
What if something didn’t come over?
This can happen. Anthropic says Claude may not keep personal details that have nothing to do with work, and it may not take in every single entry.
If something important is missing, open the backup you saved in Step 4. Copy the missing line and paste it into a chat with Claude. Tell Claude to remember it.
Your new assistant, without starting from zero
Switching tools usually feels like moving into a new house with empty rooms. You spend weeks putting everything back where it belongs.
This time, your stuff comes with you. On day one, Claude already knows who you are, what you’re working on and how you like things done. You get to skip the awkward getting-to-know-you phase and go straight to getting work done.
Want more plain-English guides like this one? Join the Pit Crew, my weekly email. One email a week, no pitch parade.
