Do I have to turn off Claude's memory to use Lodekeep?

No. They are separate. Claude's memory is built from your chats inside the Claude app; Lodekeep is a connector Claude calls to save or search memories. Both can be on.

There is a free tier with no card: one project, 250 searches and 50 captures a month, and 30-day retention. Paid plans keep memory for as long as the subscription is active.

Does Lodekeep work in Claude Code?

Yes, on the Max or Team plan. Claude Code speaks to the hosted connector directly and signs in with OAuth. The same memory is then readable from Cursor, ChatGPT, Codex and Gemini CLI.

Where do the facts about Claude on this page come from?

Anthropic's Help Center article on memory and its launch post, both read on 2026-09-15 and linked on this page. If Anthropic changes the feature, email hello@lodekeep.com and this page will be corrected.

Honest comparison

Lodekeep vs Claude’s built-in memory

Claude remembers on its own now. It is free, on by default, and for a lot of people it is enough. This page says where that is true before it says where it is not, and every fact about Claude below comes from Anthropic’s own pages, read on 2026-09-15.

Use Claude’s built-in memory if

You do your work in the Claude app and nowhere else.
You want zero setup, and you are happy for Claude to decide what is worth keeping.
You do not need to hand the same memory to Cursor, ChatGPT or a coding agent.
You would rather not pay for memory at all.

Lodekeep earns its place if

You work in Claude Code, Cursor, ChatGPT, Codex or Gemini CLI as well as Claude, and want one memory in all of them.
You want to choose what is kept, project by project, rather than have it inferred from chats.
Facts change under you, and a correction has to retire the old version instead of competing with it.
You want your memory as data you can export, and delete, on your terms.

Side by side

Where it works: Claude’s built-in memoryThe Claude app: web, desktop and mobile, plus Cowork in the cloud.; LodekeepClaude (as a connector), Claude Code, Cursor, ChatGPT, Grok, Gemini CLI and Codex. One memory, read by all of them.
Who decides what is kept: Claude’s built-in memoryClaude, automatically, from your chats. You can read, edit, delete, pause or reset it in Settings.; LodekeepYou. A memory is saved only by an explicit capture, and each project is included in or excluded from recall.
Correcting a fact: Claude’s built-in memoryEdit or delete the topic in Settings.; LodekeepSave the correction as superseding the old entry and the old version stops being recalled. Newer facts outrank stale ones, and an unknown question returns nothing rather than a guess.
Organisation: Claude’s built-in memoryA separate memory space for each project, with a project summary.; LodekeepProjects you name, each one included in or excluded from recall, and a default inbox for everything else.
Taking it with you: Claude’s built-in memoryInstructions for bringing memory details in from another AI tool. The help article does not describe an export.; LodekeepExport everything from Settings as data. Delete the account and it is erased.
Price: Claude’s built-in memoryIncluded with Claude. On by default on Free, Pro and Max; Team and Enterprise owners turn it on.; LodekeepFree tier with no card: one project, 250 searches and 50 captures a month, 30-day retention. Pro $12/mo. Coding-agent clients need Max $25/mo or Team $49/mo.
Setup: Claude’s built-in memoryNone.; LodekeepPaste one URL as a connector, or run one command in a terminal.
Coding agents: Claude’s built-in memoryAnthropic's pages describe memory for the Claude app. Claude Code keeps its own memory in project files (CLAUDE.md).; LodekeepClaude Code, Cursor, Codex and Gemini CLI read and write the same hosted memory (Max or Team).

	Claude’s built-in memory	Lodekeep
Where it works	The Claude app: web, desktop and mobile, plus Cowork in the cloud.	Claude (as a connector), Claude Code, Cursor, ChatGPT, Grok, Gemini CLI and Codex. One memory, read by all of them.
Who decides what is kept	Claude, automatically, from your chats. You can read, edit, delete, pause or reset it in Settings.	You. A memory is saved only by an explicit capture, and each project is included in or excluded from recall.
Correcting a fact	Edit or delete the topic in Settings.	Save the correction as superseding the old entry and the old version stops being recalled. Newer facts outrank stale ones, and an unknown question returns nothing rather than a guess.
Organisation	A separate memory space for each project, with a project summary.	Projects you name, each one included in or excluded from recall, and a default inbox for everything else.
Taking it with you	Instructions for bringing memory details in from another AI tool. The help article does not describe an export.	Export everything from Settings as data. Delete the account and it is erased.
Price	Included with Claude. On by default on Free, Pro and Max; Team and Enterprise owners turn it on.	Free tier with no card: one project, 250 searches and 50 captures a month, 30-day retention. Pro $12/mo. Coding-agent clients need Max $25/mo or Team $49/mo.
Setup	None.	Paste one URL as a connector, or run one command in a terminal.
Coding agents	Anthropic's pages describe memory for the Claude app. Claude Code keeps its own memory in project files (CLAUDE.md).	Claude Code, Cursor, Codex and Gemini CLI read and write the same hosted memory (Max or Team).

You can run both

They are not exclusive. Claude’s memory works from your chats; Lodekeep is a tool Claude calls when you ask it to remember or recall something, and the same tool is there in your other clients. Nothing about connecting Lodekeep changes what Claude keeps on its own.

Try the demo, no account →Connect Claude

Questions

Do I have to turn off Claude's memory to use Lodekeep?: No. They are separate. Claude's memory is built from your chats inside the Claude app; Lodekeep is a connector Claude calls to save or search memories. Both can be on.
Is Lodekeep free?: There is a free tier with no card: one project, 250 searches and 50 captures a month, and 30-day retention. Paid plans keep memory for as long as the subscription is active.
Does Lodekeep work in Claude Code?: Yes, on the Max or Team plan. Claude Code speaks to the hosted connector directly and signs in with OAuth. The same memory is then readable from Cursor, ChatGPT, Codex and Gemini CLI.
Where do the facts about Claude on this page come from?: Anthropic's Help Center article on memory and its launch post, both read on 2026-09-15 and linked on this page. If Anthropic changes the feature, email [email protected] and this page will be corrected.