How to Migrate ChatGPT Memories and Chats to Claude (Including Claude Cowork Triage)

So one of the biggest questions I get from people who want to move from, say, ChatGPT over to Claude is, what do I do with all my memories? What do I do? I like, how do I migrate it over? And there's a couple of ways to do it. I'm going to show you two. One of them is the easiest. And the second one is probably the most in-depth. Now I'm going to start off by saying there is no easy way or even a way to move over all the individual chats and then import them into, say, Claude as individual chats. But I'm going to show you the next best thing. Step 1 is if you're in Claude, come down to settings and you're going to go to, I believe it's capabilities. Then right here, you're going to see import memory from other providers. You click start import and it's going to give you a prompt that you copy and paste into, let's say, ChatGPT. So if I copy this, I come over to ChatGPT. And it'll say, export all my stored memories and any context you've learned about me, preserve words verbatim, et cetera. And you can see all the instructions in here. Now, this is then going to give you an output of— a structured output of what ChatGPT knows about you. Identity, career, projects, preferences. And what you're seeing here is a very small one because I'm using a ChatGPT account that I've only used a handful of times. This is not my primary one, and I'm doing this make sure that brief, but all you would have to do is copy this, bring it in here and paste it right into this field and then add to memory. So this adds your previous ChatGPT's memories to your Claude. Now, an interim step that I would personally do is I would paste those into, say, a Google document and I would edit them and I would remove anything you don't want Claude to remember because sometimes it gets these wrong. But then now you're bringing over contextual memories, but you're not actually bringing over any chats. If you want to bring over chats, there's a different way to do this. It's not pretty, but it is good if you want to preserve things. So inside your ChatGPT, you come down here, you go to settings, and you go to data controls, and then you're going to click export data. When you do this, it's going to confirm the export. That can easily take 24 hours depending, or even 48, depending on how much you've got before you receive it. You'll get an email and in this email it'll say, hey, your download is ready. So you're going to download that file. Now I just did that over here. And so for instance, you'll see you get a, you will get a, where is it? A zip file that looks like this. I open the zip file and I'll see a chat, conversations, export. You'll see a lot of different things. The conversations, this is the most important one, are in here. Conversations.json. We're gonna click on this. Okay. And you'll see a very long file. Now, by the way, there is not even a lot in this. So this is a very— I probably got 17 entire chats in this, and you can see how big this file is. This is not a human-readable file. It's really hard to read. I'm going to show you one of the things you can do. So when you're using Claude, and I like to use Claude Cowork for this. If you're not used to Cowork, I've got other resources on that. You can open those up. But I'm using the Claude Cowork, I'm using the Claude app here, and I'm on the Cowork feature. The very first thing I did, and let me back up, I went into here and I said, attach this to a folder. So I chose a folder, which was in my downloads. This is just me personally, and I put it in temporary. So what I'm doing is I'm pointing Claude Cowork at the folder on my computer, just like this. So this means Cowork can read from that. And this is what I did here already. And I just wanted to confirm. I said, first find the conversations.json in a folder in there. And it says, okay, I found it. And then I said, what can you do to help me understand the organizational structure or contents to decide which previous chats I might want to keep or ignore? So what Cowork is doing is it's going inside that file, it's reading that big ugly JSON file, and it's coming back with something like this. We exported 17 conversations, and this is an at-a-glance inventory, right? So there's a whole bunch of stuff here. Now, some of this stuff might be personal. I don't need it. Some of this stuff may be business-related, client-related, et cetera. Once more, I might have thousands or hundreds of these in my primary account, so it can get pretty ugly. If you're trying to do this as a human. But if you're trying to do this with AI, it gets a lot easier. But at least here I can see what these are. Now it gives me some patterns and it says, here's what I can do next. We can do a first prompt preview for each chat, pull the opening user message so you can recognize a one-line AI summary. We can keep, archive, or delete the ones. It can even make recommendations. I apply what heuristics, recency, length, repeat themes, et cetera, and give you a sortable table. Or I can group by theme, bucket them into business, marketing, career, et cetera. And I can even create an interactive HTML triage view with title, summary, full text. It can do all of these. So what do you want to do? I just confirmed there's only 17 chats and it said yes. So then the next thing I did is I just said, uh, segregate out the chats for number 8 so I could import it into a separate Claude file. So number 8 up here was just this. I just picked it randomly. So imagine that is, that's a chat I cannot lose. I want to keep that. So I didn't have it do all the recommendations and it did. It pulled it out and I can see here if I were to click on this is a, the, this is a human readable version. So it'll say, this is me. This is ChatGPT. And I think there was, this was only one question and one answer, right? It's very simple. And now this is a JSON of that chat. Now, once more, JSON is more computer readable that you can bring in. You could bring either of these into a Claude chat. You can import it. And one of the ways you would import that, it just, it automatically saved these on my folder, right? I can see I've got this and this. So let's just say I'm back in regular old, I'm back in regular old Claude Chat. I could come here. I could add this folder. Let me go into the temporary one. Oops. Where did it save it? Saved it right here. So let's just say I were to import this. Importing a previous ChatGPT file so we can continue. And I might just say, summarize what this is. Okay. Now let's say you've got projects in a, in your ChatGPT, you could create the projects over here and then pull this in. But it says, here's what it is. Boom. What do you want to do? Continue the original thread? Yes. Compare it to this. Go deeper. In essence, I might just continue my original thread. So. What you're learning here is a few things. One is how to export the memory to memory, then chat threads to chat threads, but also the power of using something like a plug, Cowork, to go in and read and potentially, and then write to those files. You not only read the original big JSON file, but it created, it segregated out the ones that I wanted to keep. Now let's just say I went back in here. Because before, I'm just going to copy this and I'll say, earlier you said you could do this, please do it. So it's going to go in and it's going to go into this little folder that I just showed you. Whoops. Where is it? It's going to go into the, yeah, the big ugly conversations folder. And then it will ideally create recommendations of what should happen next. I'm going to push pause while it does this. Okay, now it's done. So you're going to see what it did here. So it's going to say verdict. We've got 5 to keep, 10 to archive, 2 to delete. Tells you the heuristics it applied. Then it says a rundown of the 5 keeps. These 5, 2 deletes is just this and this. And then, uh, want me to do anything with these? Right? So I can see. What these look like. By the way, let me just see what it opened here. It opened up a— it created an HTML file. Look at this. This is great. I didn't even ask it to do this, but it created an HTML file that I can take a look at these. And I can see a little bit more detail of what these are about. Thank you, Claude. The next thing I can see is if I come in here, I just show this in my Finder, right? What it did, it put it in a new folder on my computer. And then these are the various chats, right? So I can run through this and I can look at it for instance. And if I click on this, I can even see it easier over on the right-hand side. So now I can make decisions and I can tell it, let's just keep, so let's say let's keep 1 through 5 only, delete the others. And then I'm going to show you what it will do. And then I'll show you how to bring this into, like, a Claude, very similar to what I just did a second ago. I'm going to come back to you real quick just to show you what this looks like. So if you want to take a closer look, a really handy way to do this. While I'm waiting, I'm going to push pause. Okay, it did it. Now let me show you what it gave me. It gave me all these files here. But what I really care about is the ones that I said to keep. There was only 5. 1, 2, 3, 4, 5. And if I take a look at each of these, this is the previous chats. Just like this. Here's the user. Here is ChatGPT. Where's the user? Here's me. Right now. So now we've exported the individual chats and we can do exactly what I showed you a second ago, which is to bring them in just like I did like this, just Upload the JSON or the Markdown. It doesn't matter. Both basically do the same thing. This is JSON, the other one's Markdown. Markdown's easier for a human to read, but JSON's easier for a computer to read. And this is how you bring in your chats. So it's really fairly simple. It's not necessarily one-click easy, but it gives you also a great way to start to work with Claude Cowork from day one. And if you have any other questions about that, There's a, uh, the link below for the Claude Co-Work Quick Start. So hopefully this is helpful.