A

Anki MCP Server

Enables AI assistants to manage Anki flashcard decks and cards through natural language, supporting ...

60 views
0 installs
Updated Feb 9, 2026
Not audited
Tools I Recommend
Enables AI assistants to manage Anki flashcard decks and cards through natural language, supporting deck creation, card additions (basic and cloze types), and review queue management.
  1. Open the FastMCP connection interface and click “Install Now”

    • Use the provided “Install Now” button for this MCP server; you will fill the ENV values directly into the FastMCP connection interface fields. (npmjs.com)
  2. Which ENV keys to provide (the FastMCP fields to fill)

    • ANKI_BASE_URL — the base URL for your Anki API (e.g., a local Anki server or remote REST endpoint).
    • ANKI_API_KEY — the API key/token used to authenticate with that Anki API (if required).
    • (Optional) API_URL — URL of an external tools API (only if you plan to use external tools).
    • (Optional) API_KEY — the key for the external tools API. (npmjs.com)
  3. If you are using AnkiConnect (the Anki desktop add-on, local Anki) — set ANKI_BASE_URL

    • Install Anki and the AnkiConnect add‑on, then set ANKI_BASE_URL to the local AnkiConnect address (e.g. http://127.0.0.1:8765 or http://localhost:8765). (github.com)
  4. If you want AnkiConnect to require/use an API key — enable it in Anki’s add‑on config

    • In Anki: Tools → Add‑ons → AnkiConnect → Config — set the apiKey (the string you choose). After enabling that config, include that same string as ANKI_API_KEY in FastMCP. You can confirm the add‑on’s requireApiKey setting via the requestPermission action. (github.com)
  5. If you are using a separate REST Anki server (not AnkiConnect) — determine base URL and how that server issues API keys

    • Set ANKI_BASE_URL to the server’s root (example: https://your-anki-server.example.com/api/v1).
    • Create or copy an API key from that server’s dashboard or developer / API keys page (typical flow: log into the server admin UI → Developer / API Keys → Create new key → copy it). Provide that key as ANKI_API_KEY. (npmjs.com)
  6. Test connectivity before pasting into FastMCP

    • For AnkiConnect (local): run a test request from the machine running Anki:
      curl localhost:8765 -X POST -d '{"action":"deckNames","version":6}'
      If you enabled an apiKey in AnkiConnect, include it in the JSON body as "key":"YOUR_KEY". (github.com)
    • For a REST server: try GET $ANKI_BASE_URL/decks using curl with the expected auth header (e.g., Authorization: Bearer <ANKI_API_KEY> or X-API-Key: <ANKI_API_KEY>) depending on that server’s docs. If the server returns your deck list, your base URL and key are correct. (npmjs.com)
  7. Fill the FastMCP connection interface fields

    • In the FastMCP connection UI, paste the exact ANKI_BASE_URL you confirmed and the ANKI_API_KEY you obtained or configured. For optional external tools, paste API_URL and API_KEY into their corresponding FastMCP fields. (npmjs.com)
  8. Save and verify in FastMCP

    • Save the connection. Use the MCP server’s test or “list decks” tool (e.g., ask the assistant “List all my Anki decks”) to verify the server responds. If it fails, re-check base URL, key value, and whether the Anki service requires the key in a header vs. request body. (npmjs.com)
  9. If you don’t need a key

    • If you’re using a local AnkiConnect installation and did not enable apiKey in the add‑on config, leave ANKI_API_KEY empty and use ANKI_BASE_URL=http://127.0.0.1:8765 (or the REST server’s base URL if applicable). (github.com)
  10. If the README or server docs mention a different auth method, follow those docs

  • The MCP server README shows the ENV names above; if your chosen Anki API provides a different key creation flow, use that service’s “API keys / developer” documentation to create the key and then paste it into FastMCP. (npmjs.com)
  1. Optional: store the ANKI_API_KEY securely
  • If FastMCP offers a secret/secure field or vault for ENV values, use it for ANKI_API_KEY and any external API_KEY so they are not shown in plain text. (npmjs.com)

If you want, tell me which Anki setup you have (Anki desktop with AnkiConnect, or a separate REST Anki server), and I’ll provide the exact example values and the precise curl commands to run.

Quick Start

View on GitHub

More for Productivity

View All →

More for API Development

View All →

Similar MCP Servers

Report Issue

Thank you! Your issue report has been submitted successfully.