The first time Claude remembered something about me without being told again, it was a small thrill....
The first time Claude remembered something about me without being told again, it was a small thrill. It opened a new chat and referenced a project I had mentioned days earlier. And then, a few minutes later, it became obvious how little it actually held onto. It knew a preference. It did not know my work.
That gap is the subject of this post. Claude has a memory now, which is a real step forward, but the built-in version is a light one. This is about what it does, where it stops, and the much bigger thing that happens when you give Claude a memory you actually control.
This is the tenth post in a beginner's series on Claude. The [first one](https://studiomeyer.io/en/blog/what-is-claude-2026) mapped out the whole tool. Here we get to the piece that, more than any model upgrade, decides whether Claude feels like a clever stranger or a colleague who knows you.
Since early 2026, every Claude account, including the free one, has a memory. It keeps a short, structured set of facts about you and your work and carries them between conversations. You can read and edit that set in the settings, and Claude tells you when it is leaning on something it remembered. If you missed the earlier post in this series on it, that is the shape of it: preferences, ongoing context, the kind of notes a good assistant jots down.
For a lot of everyday use, that is genuinely enough. It stops you re-introducing yourself every morning. It is a real improvement over the years when every chat started cold.
Now the limits, because they matter. The built-in memory is light on purpose. It holds preferences and light context, not the deep, specific knowledge of your business that would make it a true partner. It lives inside Claude and only Claude, so the moment you move to another tool, that memory does not come with you. And it is not really yours to shape, move, or build on. It is a convenience feature, and a good one, but it is a long way from a second brain.
You feel this the moment you want more than convenience. You want it to remember every decision you made on a project and why. The names and quirks of your twelve clients. The reasons you rejected an approach three months ago so you do not relitigate it. The built-in memory was not built to carry that, and it will not.
The next step, and the one that changes how the whole thing feels, is an external memory that you own. Instead of a light set of notes locked inside one app, you connect a dedicated memory that any tool can read from and write to. It accumulates the things that actually matter over time, your decisions, your context, your history, and it follows you across whatever you happen to be using. The way you wire that up is through connectors, which is the subject of the next post in this series, so I will not go deep here. The point for now is that this category exists and it is the difference between an assistant