TL;DR: To transfer your ChatGPT memory to Claude, ask ChatGPT to list everything it remembers about you in one block, then paste that into Claude's built-in memory import (Settings, Memory, "Start import"). It takes a few minutes, but it is a one-time copy: the next change, and the next AI you try, means doing it again. Or paste the same list into UseMyContext.ai once: every AI you connect reads the same live version, and only you can change it.
How do you see what ChatGPT remembers about you?
ChatGPT keeps two things that matter here. Its memory is what it has picked up from your chats; you can review and delete it under Settings, Personalization, Memory, as OpenAI's memory FAQ describes. Your custom instructions are what you wrote yourself, in the same Personalization area (OpenAI help).
The quickest way to get the memory out as text is to ask for it, in ChatGPT. Anthropic's suggested prompt starts like this (Claude shows the full version when you click Start import):
I'm moving to another service and need to export my data. List every memory you have stored about me, as well as any context you've learned about me from past conversations. Output everything in a single code block so I can easily copy it. Format each entry as: [date saved, if available] - memory content.Then copy your custom instructions from the Personalization page; that is the easiest route. They are often the most useful part, because you wrote them on purpose.
How do you import it into Claude?
Claude has a real, built-in import, and it is the right first move. Per Claude's help center:
- In Claude, open Settings, then Memory, and choose "Start import". It shows the full prompt and a text box.
- Paste what ChatGPT gave you, plus your custom instructions.
- Click "Add to memory". Claude turns it into memory entries you can see and edit.
On older versions of the settings, imported memory can take up to 24 hours to appear.
It works on the Free, Pro, Max and Team plans, on the web and in Claude Desktop. Two honest caveats from Anthropic's own page: the import is experimental and "Claude may not always successfully incorporate imported memories", and it may not keep personal details unrelated to work. Check what landed afterwards.
Keep the list you pasted. The same text works as a starting point for one context every AI reads, below.
What is worth moving from ChatGPT to Claude?
Not everything ChatGPT remembers is worth keeping. These five are:
- Who you are. Your role, your field, what your work actually involves.
- How you want answers. Length, tone, format, the language you write in.
- What you are working on. Current projects, goals and deadlines.
- Your tools. The software, stack or methods you use every day.
- Your standing corrections. The "stop doing X" rules you taught it the hard way.
Leave out stale projects, one-off trivia and anything you would not want a second company to hold. What is left is also a good first profile for UseMyContext.
What is the catch?
It is a copy, not a connection. The moment you change jobs, start a project or fix a preference, Claude and ChatGPT disagree again, and nothing syncs them.
And every new AI asks for the same ritual. Gemini now has its own memory import too. The next tool, the second account: each one is another paste, and another copy drifting out of date.
What does UseMyContext add that Claude's import can't?
Claude's import is a good way to get into Claude. UseMyContext is built for the case where you use more than one AI and want them to agree.
| Claude's import | UseMyContext | |
|---|---|---|
| Copy or live | A one-time copy into Claude's memory | One source every connected AI reads when it needs it |
| Who changes it | Claude updates it from your chats; you can view and edit it | An AI can only suggest an update; nothing changes until you approve it |
| Which AIs | Claude | Claude and ChatGPT, plus coding tools like Gemini CLI and Cursor |
| Your documents | Text memories | Your CV, notes and PDFs too, which a connected AI can read and quote |
| Who read it | Used inside Claude | An activity feed shows which AI read your profile or opened a file |
How do you move it once, to UseMyContext?
UseMyContext.ai stores your profile, facts and files once, for any AI that speaks MCP, the open standard for connecting tools to AI. Moving in uses the same ChatGPT list:
- Sign up at usemycontext.ai with your email; you sign in with a code, no password. The free plan includes 2 projects and 20 MB.
- Name your profile and claim your @handle, then at "Seed", choose "Write or paste a blurb". Paste the ChatGPT list plus your custom instructions and click "Seed my context". UseMyContext turns it into facts you can edit.
- Connect your AI. Claude installs from the Claude Directory and ChatGPT from its app listing; you sign in with your email and the code we send. You can disconnect any one of them later from the Connect page.
Use Claude's import today if you are only switching. If you plan to keep using more than one AI, move your context once, to UseMyContext.
FAQ
Does Claude have a memory import from ChatGPT?
Yes. In Claude's Settings, under Memory, "Start import" gives you a prompt to run in ChatGPT and a box to paste the result into. Anthropic labels it experimental.
Does ChatGPT's data export include my memories?
Not your memories. The export covers your chats and your custom instructions, but saved memories have no export button. Asking ChatGPT to list them, as above, is the reliable way to get them out.
Can Claude and ChatGPT share one memory?
Not natively; each keeps its own. A context layer like UseMyContext gives both the same source to read over MCP, so an update in one place reaches both.