What features does Anki MCP server provide?

Q: What features does Anki MCP server provide?

Card creation and update: Create new flashcards, update existing notes, and support bulk additions using HTML-formatted content. Supports input for fields, modelName, deckName, and tags, with note updates via update_note_fields and bulk operations via add_bulk/bulk_update_notes.. Audio-enabled card operations: Create cards with automatically generated audio via Azure TTS and update audio on existing cards. Supports specifying source and target audio fields and language selection.. Quiz-driven updates: After quizzing, use update_cards to mark cards as answered and adjust their ease, enabling adaptive study workflows.. Bulk operations for speed: Efficient bulk handling of cards and notes with add_bulk and bulk_update_notes to create or update multiple items in a single operation.. Card search and retrieval: Find cards with raw Anki queries (find_cards), and retrieve due or new cards (get_due_cards, get_new_cards) and deck names (get_deck_names).. Model and deck management: Inspect and manage note types/models (get_model_names, get_model_details), create models, and modify fields (add_note_type_field, remove_note_type_field, rename_note_type_field, reposition_note_type_field) along with templates and styling updates.. Deck and model metadata queries: Retrieve deck-wide model usage information (get_deck_model_info) to determine whether a deck uses a single or multiple models, or if it’s empty or non-existent.. Deployment and integration tooling: Run locally via npx or source, build and package into a Claude Desktop MCP bundle (.mcpb), and integrate with Cursor workflows for automatic server startup and local automation.

Anki MCP Server 2025: Features & Setup Guide

Overview

anki-mcp is an MCP server for Anki that enables programmatic control of flashcards, decks, and review workflows via the Model Context Protocol. It exposes a suite of tools to create and update cards and notes, search and filter content, and inspect deck and model information. The server supports both single-card operations and bulk updates for efficiency, as well as advanced features like audio generation via Azure TTS and automatic media handling through the Anki collection.media directory. It can run locally (via npx or from source) and is designed to work with local AnkiConnect installations; instructions cover environment setup, including optional Azure and media configurations. The tooling includes functions for retrieving due and new cards, finding cards with raw Anki queries, creating decks, and managing note types and templates. The README also documents packaging as an MCP bundle (.mcpb) for Claude Desktop and integration with Cursor-based workflows, enabling one-click deployments and automated server startup.

Features

Card creation and update

Create new flashcards, update existing notes, and support bulk additions using HTML-formatted content. Supports input for fields, modelName, deckName, and tags, with note updates via update_note_fields and bulk operations via add_bulk/bulk_update_notes.

Audio-enabled card operations

Create cards with automatically generated audio via Azure TTS and update audio on existing cards. Supports specifying source and target audio fields and language selection.

Quiz-driven updates

After quizzing, use update_cards to mark cards as answered and adjust their ease, enabling adaptive study workflows.

Bulk operations for speed

Efficient bulk handling of cards and notes with add_bulk and bulk_update_notes to create or update multiple items in a single operation.

Card search and retrieval

Find cards with raw Anki queries (find_cards), and retrieve due or new cards (get_due_cards, get_new_cards) and deck names (get_deck_names).

Model and deck management

Inspect and manage note types/models (get_model_names, get_model_details), create models, and modify fields (add_note_type_field, remove_note_type_field, rename_note_type_field, reposition_note_type_field) along with templates and styling updates.

Deck and model metadata queries

Retrieve deck-wide model usage information (get_deck_model_info) to determine whether a deck uses a single or multiple models, or if it’s empty or non-existent.

Deployment and integration tooling

Run locally via npx or source, build and package into a Claude Desktop MCP bundle (.mcpb), and integrate with Cursor workflows for automatic server startup and local automation.