Does Claude Remember Past Conversations?

Yes. Claude carries context from your past conversations into new ones, within your own account and only while memory is on. As of 2026-09-14, the Claude Help Center states that memory "is on by default for Free, Pro, and Max plans on the web, Claude Desktop, and Claude Mobile," while on Team and Enterprise plans it "is off by default and can be turned on by an owner."

How that memory is built has just changed, and the difference is worth knowing before you go looking for a setting. In the current experience Claude "saves memory as a set of individual topics as you chat, rather than summarizing conversations after they end." Older accounts still on the legacy experience get the previous behavior: a synthesis of your chat history that, per the same page, "is updated every 24 hours." Memory is also scoped per project, and Claude is the one major assistant that lets you move your memory out as plain text. This article sets out what Claude retains, where each control lives, and where its memory stops. Every vendor fact below was checked against Anthropic's own documentation on 2026-09-14; these pages were last revised within the past week, so treat the exact labels as a snapshot.

Where Claude's memory lives

Before anything else, find out which experience your account is on, because the settings path differs.

The Claude Help Center gives a one-step test, quoted here as of 2026-09-14: "If you see Settings > Memory, you're using the new memory experience." If instead you find Memory sitting inside Settings → Capabilities, you are on the legacy experience and the older behavior described below applies to you.

Anthropic describes the migration as in progress. Per the import and export documentation, the new experience "will be the default for new users, and users on free, Pro and Max plans will be migrated"; Team and Enterprise accounts "will stay on the legacy experience" in the interim. So two accounts on the same plan can legitimately show different screens today.

On the new experience the master control is Settings → Memory, where the relevant toggle is "Generate memory from chats." That same page is where you see what Claude holds, edit it, and start an import.

What Claude actually remembers

The unit of memory changed with the new experience, and it changes what you should expect to see.

In the current experience, per the Claude Help Center, Claude "saves memory as a set of individual topics as you chat, rather than summarizing conversations after they end." Memory accumulates continuously as discrete entries rather than arriving as a digest after the fact. On the legacy experience the same page describes the older model: Claude "will automatically summarize your conversations and create a synthesis of key insights across your chat history," and that synthesis "is updated every 24 hours." If you have read elsewhere that Claude's memory refreshes daily, that is the legacy behavior specifically.

What Claude retains is oriented toward work. The documentation describes memory focused on your role, projects, and professional context, your communication preferences and working style, your technical preferences and coding style, and the details of ongoing work. The import and export documentation states the design goal plainly: memory "is designed to focus on work-related topics," and Claude "may not retain imported personal details unrelated to work." It is built to be a useful collaborator rather than a general diary of your life.

If you use projects, memory is scoped to them. Anthropic states that each project has "its own separate memory space and dedicated project summary, so the context within each of your projects is focused, relevant, and separate from other projects or non-project chats," per the Claude Help Center. Work you do in one project does not bleed into another.

How to see and control it

Every control lives under settings, and Anthropic documents the paths for both experiences.

On the new experience, open Settings → Memory to see what Claude remembers about you and to turn "Generate memory from chats" on or off. On the legacy experience the same controls sit under Settings → Capabilities. You can add to memory by telling Claude in conversation what to remember, or, per the import and export documentation, by using the "Tell Claude what to change or remove" box on the memory page.

There are two distinct ways to turn memory off, and the difference matters. Pause memory keeps what Claude already holds but stops it from acting: per the Claude Help Center, Claude "keeps its existing memory ... but won't use memory or make new memories." Reset memory "permanently deletes all memories including project memories," and Anthropic states that once you confirm it, "this cannot be undone." Pause is reversible; reset is not.

Memory portability: export and import

One capability sets Claude apart from most assistants: your memory is not locked in. Anthropic lets you move it as plain text, in both directions.

To get your memory out, the import and export documentation describes asking Claude directly in a chat to "Write out your memories of me verbatim, exactly as they appear in your memory," then saving the result as a backup or carrying it to another service. There is no file-download button; the export is a conversation you copy.

To bring memory in, go to Settings → Memory and select "Start import" on the new experience, or Settings → Capabilities on the legacy one. Anthropic supplies a suggested prompt for extracting your memory from another provider, covering instructions about response style, personal details, projects and goals, the tools and languages you use, and corrections you have made. Paste the result and click "Add to memory," and Claude "will extract key information and store it as individual memory entries."

Two caveats, both stated by Anthropic as of 2026-09-14. Imports "are experimental and still in active development," and Claude "may not always successfully incorporate imported memories" — treat an import as a head start, not a guaranteed transfer. And availability is narrower than memory itself: imports "are available for Free, Pro, Max, and Team plans on the web and Claude Desktop," which excludes Enterprise and the mobile apps.

Chat search vs. memory

Claude offers two related but distinct capabilities, and it is easy to conflate them.

Chat search lets Claude search your previous conversations to find and reference specific information on request. Memory is the standing context Claude carries into new chats on its own. The Claude Help Center frames them as related but separate: search is a lookup you invoke, while memory shapes new conversations without you asking. Turning one off does not turn off the other.

What Claude's memory does not do

Memory has real boundaries, and knowing them prevents both false expectations and false worries.

It is bound to your account. Memory is a feature of Claude tied to your own account and its settings. The export path above gives you a copy of the text, but the memory itself does not follow you into another company's assistant on its own.

Projects are isolated from each other. As noted above, each project has its own separate memory space and dedicated summary. Memory built inside one project stays within that project's context.

Incognito chats deliberately remember nothing. For a conversation you do not want retained, Claude offers incognito chats, available on all plans. Per the incognito chat documentation, they "aren't saved to your chat history or to Claude's memory," starting one "won't use Claude's existing memory," and they "will not be included in future memory entries." They are still retained for 30 days by default, or longer under an organization's custom retention setting. You start one via the ghost icon when beginning a chat outside a project, and an incognito chat cannot later be converted to a regular one.

Memory vs. context window

The most common confusion is between memory and the context window, which are distinct.

The context window is the working span of text the model can see within a single conversation: your current messages plus the model's replies, up to a size limit. It is short-term and local. When a conversation grows long enough, earlier turns fall outside that window and the model can no longer see them directly. This is not memory in the feature sense; it is the model's immediate attention, and it resets when you start a new chat.

Memory is the persistent layer that survives across separate conversations. Claude's stored topics are pulled into new chats as relevant context, which is why Claude can carry your working style or project context into a brand-new conversation even though that conversation's context window started empty. A concise way to hold the distinction: the context window is what Claude can read in front of it right now, and memory is what it carries in from before. Losing track of something mid-conversation is usually a context-window limit, not a memory failure.

FAQ

Does Claude remember me between separate conversations? Yes, if memory is on. As of 2026-09-14 the Claude Help Center states memory is on by default for Free, Pro, and Max plans on the web, Claude Desktop, and Claude Mobile, and off by default on Team and Enterprise until an owner turns it on. If memory is off, each conversation starts blank.

Does Claude's memory still update every 24 hours? Only on the legacy experience. Per the Claude Help Center, the legacy synthesis "is updated every 24 hours," while the current experience saves memory "as a set of individual topics as you chat, rather than summarizing conversations after they end." Check whether your settings show Memory or Capabilities to know which applies.

How do I make Claude forget something? Open Settings → Memory on the new experience, or Settings → Capabilities on the legacy one, and remove the detail. To clear everything, use Reset memory, which "permanently deletes all memories including project memories" and cannot be undone, per the Claude Help Center. If you only want a temporary stop without deleting anything, use Pause memory instead.

Can I move my Claude memory somewhere else, or bring memory in? Yes. You can ask Claude in chat to write out its memories verbatim and save that text, and you can import memory exported from another provider via Start import, per the import and export documentation. Anthropic notes imports are experimental and available on Free, Pro, Max, and Team plans on the web and Claude Desktop.

Is memory the same as Claude's context window? No. The context window is the short-term text the model can see within one conversation, and it resets each chat. Memory is the persistent layer that survives across conversations. Forgetting something mid-chat is usually a context limit, not a memory setting.