Why ChatGPT and Claude don't share memory
Both apps can remember you. ChatGPT keeps saved memories and can refer back to your past chats; you manage them under Settings, Personalization, Memory (OpenAI's help page). Claude builds a memory from your chats, with a separate memory for each project; you manage it under Settings, Memory (Anthropic's help page).
Each memory lives in that company's account, on that company's servers, and there is no link between the two. So the preferences ChatGPT learned over months mean nothing to Claude on the first day, and the reverse. People describe it the same way again and again: re-explaining who they are, what they're working on and how they like answers, every time they switch.
There are three ways around it. They differ in effort and in one thing that matters a lot: whether they keep working after the first day.
Option 1: copy and paste
The simplest way costs nothing:
- In ChatGPT, ask: "List everything you remember about me, as short bullet points."
- Read the list and remove anything you don't want to carry over.
- In Claude, start a chat, paste the list and say "Please remember these about me."
It works for a handful of facts. The catch is that it is a snapshot. Anything you tell ChatGPT tomorrow stays in ChatGPT until you copy again, and the two memories drift apart from there.
Option 2: Claude's memory import
Claude has a tool built for this move (Anthropic's help page). It is on the Free, Pro, Max and Team plans, on the web and in Claude Desktop:
- In Claude, open Settings, then Memory, and choose Start import. Claude gives you a prompt to copy.
- Paste that prompt into ChatGPT. It writes out what it remembers about you.
- Copy ChatGPT's answer back into Claude's box and choose Add to memory.
It is quick, but Anthropic is clear about the limits: the import is experimental, it can take up to 24 hours to show up, and Claude keeps mainly work-related details, so some personal ones may not carry over. It also only goes one way, once. OpenAI's memory help page describes no import from other apps, and a fact you tell ChatGPT after the import never reaches Claude.
Which option to choose
| Option | Effort | After the first day | Where it's kept |
|---|---|---|---|
| Copy and paste | A few minutes, every time | Drifts apart until you copy again | In each app's account |
| Claude's memory import | A few minutes, once | Only ChatGPT to Claude, only what was there at the time | In Claude's account |
| One shared memory | Connect each app once | Both apps save and recall, both ways | Wherever that memory runs; with Zaaheen, on your computer |
We tested a shared memory ourselves. Five AI apps on one Windows computer were connected to the same Zaaheen memory, and we asked eight questions about the person, such as their favourite tea, their favourite city and when they like meetings, mostly in a different app from the one that saved the answer.
Every app that reached the memory gave the right answer. Half the time the app checked by itself; the other times it answered from its own memory first and needed "check Zaaheen". That is why the line above, in each app's instructions, matters.
If you only use the two apps now and then, copying a short list once may be all you need. If you're moving from ChatGPT to Claude for good, Claude's import is made for that. If you use both, every week, a shared memory is the one that doesn't go stale.