Overview - supermemory | Memory API for the AI era

Supermemory MCP gives every MCP-compatible assistant a shared memory layer, so technical and non-technical teams can collaborate with AI using the same authorized context. Connect once to search decisions, save knowledge, upload source material, and explore relationships from the AI tools your team already uses. Engineering and research teams can carry project context across assistants, while finance, legal, medical, and operations teams can work from their own source material. Ask naturally. The assistant selects the right Supermemory tool without requiring code or tool names. See the tools reference for exact inputs and results.

Connect

Add this remote MCP server to any compatible client:

https://mcp.supermemory.ai/mcp

After you connect, your client opens Supermemory in a browser. Sign in or create an account, then choose which spaces the client can access. Supermemory uses OAuth, so no API key is required. For client-specific instructions, see Setup and Usage or the Claude Desktop guide.

Manual JSON configuration

MCP clients that use JSON configuration generally accept this shape:

{
  "mcpServers": {
    "supermemory": {
      "url": "https://mcp.supermemory.ai/mcp"
    }
  }
}

How spaces work

A space keeps a team’s documents, memories, and profile context focused, so AI retrieves the right knowledge without mixing unrelated work. Use separate spaces for an engineering launch, research project, finance workflow, legal matter, medical knowledge base, or any other shared initiative. Teammates collaborate within the spaces they are allowed to read or write. Name a space to use it for one request without changing your active space. Otherwise, Supermemory uses your active space or account default. Ask to switch spaces when you want future requests to use a different active space.

Tools

Your assistant chooses these tools automatically. Use this table when you need the exact inputs and results.

Tool	Use it for	Inputs	Result
`search_memory`	Semantic recall from one space	`query` (required), `containerTag`	Matching memories
`get_profile`	Stable and recent profile context for one space	`containerTag`	Profile facts
`add_memory`	Save information or forget outdated information	`content` (required), `action` (`save` or `forget`), `containerTag`	Save or forget confirmation
`list_documents`	Browse stored source documents and their summaries	`page`, `limit`, `containerTag`	Document IDs, titles, types, status, dates, and summaries
`get_document`	Read the available content of one document	`documentId` (required)	Document metadata, summary, and available content
`list_memories`	Browse recent extracted memory entries and their source document IDs	`page`, `limit`, `containerTag`	Memory IDs, text, versions, and source document IDs
`list_spaces`	List accessible spaces and resolve a space name to its key	None	Formatted list plus structured `spaces` and `count` fields
`who_am_i`	Inspect the authenticated account, permissions, scope, and active space	None	Account and access context

Search

search_memory accepts a natural-language query and returns semantically relevant memories. It does not include the space profile. Use get_profile when the assistant needs stable facts or recent context that search did not return. Use the retrieval tools for different questions:

Use search_memory to answer a question from remembered context.
Use get_profile for who-the-user-is, preferences, and recent context.
Use list_documents to discover stored sources, then get_document to read one source in full.
Use list_memories to inspect the extracted memory entries themselves, including their IDs and source document IDs.

list_documents and list_memories default to 10 results per page and accept up to 50.

Save or forget

add_memory saves the supplied content by default. Set action to forget when a fact is outdated or should be removed. There is no separate forget tool. If the content is already final, the assistant should use add_memory. If you want to review, edit, or choose a space before saving, it should open the guided-save widget instead.

Access control

who_am_i returns the current identity, role, access type, granted scope, and active space. list_spaces returns only spaces the authenticated account can access, with names, keys, document and memory counts, and recent activity.

Interactive widgets

Clients that support MCP Apps can open these Supermemory widgets directly inside the conversation.

Tool	What it opens
`select-space`	A searchable space picker
`guided-save`	An editable memory form with a space selector
`upload-file`	A local file picker and upload form
`memory-graph`	An interactive graph of documents and memories

Select a space

select-space opens a searchable space picker and changes the active space for future Supermemory actions. A one-off request in another space does not change the active space.

Guided save

guided-save opens a draft with editable memory content and a writable-space selector. The assistant can prefill the draft, but nothing is saved until you submit it.

File upload

upload-file opens a local file picker and uploads one file at a time to a writable space. Supported types include text, Markdown, PDF, Word, CSV, common images, MP3, WAV, M4A, MP4, and WebM.

Memory graph

memory-graph renders the selected space as an interactive graph of source documents and extracted memories. When no space is named, the server automatically uses the active space or account default. If you name a specific space, the assistant calls list_spaces to resolve its name to a space key, then opens that space’s graph.

Resources and context prompt

Some MCP clients also expose resources and prompts:

Kind	Name or URI	What it returns
Resource	`supermemory://profile`	Stable and recent profile context for the active space
Resource	`supermemory://spaces`	A compact list of accessible spaces with the active space marked
Prompt	`context`	A ready-to-attach context message for the active space, plus up to three recently active spaces

The context prompt takes no arguments. It returns profile context for the active space and up to three recently active spaces.

Setup and Usage

Configure Supermemory in supported MCP clients.

Supermemory MCP on GitHub

View the source code.

​Connect

​How spaces work

​Tools

​Search

​Save or forget

​Access control

​Interactive widgets

​Select a space

​Guided save

​File upload

​Memory graph

​Resources and context prompt