raine/anki-llm

What it solves

It automates the creation, improvement, and bulk-processing of Anki flashcards. It removes the manual effort of drafting cards, adding audio, and managing note templates by integrating Large Language Models (LLMs) and text-to-speech (TTS) directly into the Anki workflow.

How it works

The tool operates as a scriptable CLI that connects to Anki via the AnkiConnect add-on. It allows users to:

• Generate and Process: Create cards interactively through a terminal UI or bulk-process existing decks and files (CSV/YAML) using LLMs.
• Manage Workspaces: Store prompts, note types, and settings in plain files for versioning and sharing.
• Integrate Audio: Use various TTS providers to fill audio fields in bulk or during card generation.
• Edit Templates: Pull HTML/CSS card templates into a local editor and push changes back to Anki.
• Connect Models: Interface with multiple providers including OpenAI, Gemini, DeepSeek, xAI, OpenRouter, and Ollama.

Who it’s for

Language learners and students who use Anki for spaced repetition and want to leverage AI to accelerate card creation and enhance their study materials with high-quality audio and structured content.

Highlights

• Interactive Generation TUI: Compare, edit, and regenerate card candidates before importing them.
• Direct Deck Updates: Update decks in-place with support for previews, snapshots, and rollbacks.
• Agent-Ready: Provides a JSON-oriented CLI (anki-llm query) to give AI agents structured access to Anki collections.
• Flexible Model Support: Compatible with any OpenAI-compatible chat completions endpoint.