home / mcp / ankimcp mcp server

AnkiMCP MCP Server

Provides a local MCP endpoint for Anki to enable AI-assisted study and collection management.

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "ankimcp-anki-mcp-server-addon": {
      "url": "http://127.0.0.1:3141/"
    }
  }
}

You expose your Anki collection to AI assistants via a local MCP server that runs inside Anki. This enables AI-powered study sessions, card creation, and collection management by connecting your client to a nearby endpoint at localhost:3141.

How to use

Connect an MCP client to your local AnkiMCP Server to begin interactive AI-assisted study and management. The server runs in the background and communicates over HTTP, so your client can issue commands to sync decks, manage notes, fetch due cards, present cards for review, and apply styling to note types. You simply point your MCP client to the local URL and start using the features you need.

How to install

Prerequisites: ensure Anki is installed and you have access to a running Python environment (Python 3.13 is supported). You do not need to install anything extra for the MCP server if you are using the addon version inside Anki.

Install the AnkiMCP Server addon from AnkiWeb or via GitHub Releases. After installation, restart Anki to complete setup. The server will start automatically when Anki opens.

Connect the Claude Desktop MCP client (or any compatible MCP client) to the local server by configuring the client to point to the local URL: http://127.0.0.1:3141/.

Configuration

You can adjust how the local MCP server runs and its networking details from Anki’s Add-ons configuration for AnkiMCP Server. Change the HTTP mode, port, host, and startup behavior to fit your setup.

{
  "mode": "http",
  "http_port": 3141,
  "http_host": "127.0.0.1",
  "auto_connect_on_startup": true
}

Connect with Claude Desktop

Add the Anki MCP server to Claude Desktop by editing the Claude configuration file and including the local server URL under mcpServers. This enables Claude to access your Anki collection during sessions.

{
  "mcpServers": {
    "anki": {
      "url": "http://127.0.0.1:3141/"
    }
  }
}

Available tools

sync

Synchronize collection with AnkiWeb or ensure the local state is up to date.

list_decks

List all decks in your Anki collection.

create_deck

Create a new deck in your collection.

find_notes

Search notes using Anki search syntax.

notes_info

Get detailed information about notes in your collection.

add_note

Add a new note to a deck.

update_note_fields

Update fields for existing notes.

delete_notes

Delete notes from the collection.

get_due_cards

Retrieve cards due for review.

present_card

Fetch the content of a card for review.

rate_card

Rate a card after review (Again/Hard/Good/Easy).

model_names

List available note types (models).

model_field_names

Get field names for a note type.

model_styling

Get CSS styling for a note type.

update_model_styling

Update CSS styling for a note type.

create_model

Create a new note type.

store_media_file

Store a media file (image/audio) in the collection.

get_media_files_names

List media files that match a pattern.

delete_media_file

Delete a media file.

gui_browse

Open the card browser with a search query.

gui_add_cards

Open the Add Cards dialog.

gui_edit_note

Open the note editor for a specific note.

gui_current_card

Get info about the currently displayed card.

gui_show_question

Show the question side of the current card.

gui_show_answer

Show the answer side of the current card.

gui_select_card

Select a specific card in the reviewer.

gui_deck_browser

Navigate to the deck browser.

gui_undo

Undo the last operation.

system_info

Report Anki version and system information.

review_session

Run a guided review session workflow.