Guide · Resources

Anki to Deckbase migration playbook

A low-risk migration workflow for APKG and CSV paths, with explicit quality gates so import speed does not damage retention quality.

Deckbase Editorial Team8 min read

Who this playbook is for

This guide is for learners who already have active Anki decks and want a structured move to Deckbase without breaking daily review consistency. If you are searching for anki import csv, this is the implementation path that separates quick transfer from quality stabilization.

Pre-migration checklist

1
Back up every source deck before first import.
2
List active decks from the last 30 days and migrate those first.
3
Define pass/fail quality gates before scaling beyond one deck.
4
Set a daily review floor (time-based or cards-based) to protect habit continuity.

Path A: APKG-first migration

Use APKG when speed and structure preservation matter most. This path is ideal for a first migration wave where the goal is continuity, not perfect cleanup.

Import one active deck and inspect a 30-card sample immediately.
Fix high-impact issues only: unreadable prompts, broken media, obvious duplicates.
Resume review in the same day instead of pausing for large cosmetic edits.

Path B: CSV-first migration

Use CSV when field-level normalization is required. CSV takes longer up front, but it is easier to enforce naming consistency, template rules, and deduplication at scale.

1
Normalize field names and term formatting before import.
2
Create one canonical mapping for prompt, answer, tags, and context.
3
Import a pilot batch first and pass all quality gates before full rollout.

Field mapping matrix

Field	Purpose	Quality rule	Operational note
Front	Prompt text	One concept only	Split overloaded cards before import
Back	Expected answer	Direct answer first	Keep long explanations in notes
Tags	Topic grouping	Consistent taxonomy	Use domain/chapter naming
Extra / Notes	Context support	Short and optional	Avoid embedding answer clues
Media	Visual/audio context	Attached and readable	Verify broken references in sample

Front

Purpose: Prompt text
Quality rule: One concept only
Operational note: Split overloaded cards before import

Back

Purpose: Expected answer
Quality rule: Direct answer first
Operational note: Keep long explanations in notes

Batch quality gates

Run gates before every new import batch. If one gate fails, pause new imports and repair card quality first. This keeps problems small and prevents backlog debt.

Gate	Pass threshold	Fail signal
Comprehension	>=90% sampled cards understandable	Prompt format is ambiguous or noisy
Duplicate rate	<3% duplicate front prompts	Batch has repeated definitions or headings
Session load	Session time increase <15% WoW	Imported cards are too verbose
Lapse trend	Stable or improving by week 2	Prompt quality is low for core topics

Comprehension

Pass threshold: >=90% sampled cards understandable
Fail signal: Prompt format is ambiguous or noisy

Duplicate rate

Pass threshold: <3% duplicate front prompts
Fail signal: Batch has repeated definitions or headings

Session load

Pass threshold: Session time increase <15% WoW
Fail signal: Imported cards are too verbose

Lapse trend

Pass threshold: Stable or improving by week 2
Fail signal: Prompt quality is low for core topics

Anki-specific edge cases to handle before scale

Most migration regressions come from edge cases that look small in a sample and become expensive at volume. Check cloze rendering, suspended card states, leech behavior, and media references before importing additional decks.

Edge case	Common failure	Mitigation
Cloze cards	Missing cloze markers after import	Validate sample cloze rendering before full migration
Suspended cards	Unexpected active review load	Export suspension state and re-apply after import
Leech cards	Chronic failures pollute queue	Tag leeches and run rewrite/archive policy
Media refs	Broken image or audio pointers	Run broken-media scan on first 100 cards
Deck options	Interval behavior mismatch	Run 14-day metrics before scaling additional decks

Cloze cards

Common failure: Missing cloze markers after import
Mitigation: Validate sample cloze rendering before full migration

Suspended cards

Common failure: Unexpected active review load
Mitigation: Export suspension state and re-apply after import

Leech cards

Common failure: Chronic failures pollute queue
Mitigation: Tag leeches and run rewrite/archive policy

Media refs

Common failure: Broken image or audio pointers
Mitigation: Run broken-media scan on first 100 cards

Deck options

Common failure: Interval behavior mismatch
Mitigation: Run 14-day metrics before scaling additional decks

CSV import contract for repeatable migrations

If you run recurring imports, define a fixed CSV contract and version it. This reduces schema drift across operators and makes QA automation easier. A practical baseline is a six-column contract with identity, prompt, answer, context, taxonomy, and source fields.

Column	Role	Constraint
note_id	Stable row identity	Prevents accidental duplicate ingestion
front_text	Prompt field	Normalize punctuation and unicode variants
back_text	Answer field	Keep direct answer in first sentence
context_hint	Optional disambiguation	Use short domain context only
tags_pipe	Taxonomy grouping	Use pipe-delimited canonical tags
source_ref	Traceability	Store chapter/page for QA backtracking

note_id

Role: Stable row identity
Constraint: Prevents accidental duplicate ingestion

front_text

Role: Prompt field
Constraint: Normalize punctuation and unicode variants

back_text

Role: Answer field
Constraint: Keep direct answer in first sentence

context_hint

Role: Optional disambiguation
Constraint: Use short domain context only

tags_pipe

Role: Taxonomy grouping
Constraint: Use pipe-delimited canonical tags

source_ref

Role: Traceability
Constraint: Store chapter/page for QA backtracking

Keep UTF-8 encoding, enforce normalized line breaks, and avoid hidden spreadsheet formulas in export cells. Those small constraints prevent silent import corruption.

14-day stabilization protocol

1
Days 1-3: migrate one active deck and restart daily reviews immediately.
2
Days 4-7: repair top failing cards and deduplicate front prompts.
3
Days 8-10: standardize tags/templates and validate session-time trend.
4
Days 11-14: scale migration only if completion and lapse metrics remain healthy.

Scenario-based migration guidance

Scenario	Primary objective	Recommended approach
Exam sprint (8-12 weeks)	Keep daily completion stable	Import only active decks, pause archive migration
Long-term language study	Preserve sentence context	CSV path with strict field normalization
Team migration	Template consistency	Publish one canonical mapping before bulk import
Archive cleanup	Retire low-yield decks	Migrate by topic and drop stale cards early

Exam sprint (8-12 weeks)

Primary objective: Keep daily completion stable
Recommended approach: Import only active decks, pause archive migration

Long-term language study

Primary objective: Preserve sentence context
Recommended approach: CSV path with strict field normalization

Team migration

Primary objective: Template consistency
Recommended approach: Publish one canonical mapping before bulk import

Archive cleanup

Primary objective: Retire low-yield decks
Recommended approach: Migrate by topic and drop stale cards early

FAQ

Should I use APKG or CSV first?

Use APKG first for fast structure transfer. Use CSV when you need strict field cleanup or template normalization before scale.

How many decks should I migrate at once?

Start with one active deck and stabilize for 7-14 days. Expand only when completion, session time, and lapse trend stay healthy.

Can I keep Anki and Deckbase in parallel during migration?

Yes. Many learners keep legacy decks in Anki while creating net-new cards in Deckbase, then consolidate once quality controls are stable.

What if my cards don't import correctly?

Common issues include encoding problems (use UTF-8), missing required fields, and template mismatches. Export to CSV, fix the issues in a spreadsheet, and re-import. The playbook includes a CSV contract for repeatable migrations.

Can I rollback if migration doesn't work?

Yes. Before migration, export your Anki deck as backup (.apkg). Deckbase won't delete your original Anki data. IfDeckbase doesn't meet your needs, you can always return to Anki without losing progress.

What should I do in the first week after migration?

Focus on stabilization: repair failing cards, deduplicate prompts, and establish your daily review routine. Don't add new decks until your primary deck is stable. Monitor session time and lapse rates closely.

Will I lose my review history when migrating?

Basic card content transfers with APKG import. Full review history (learning state, intervals) may reset depending on import method. The FSRS algorithm in Deckbase will rebuild optimal intervals based on your new review performance.

First week after migration: success checklist

The first seven days determine whether your migration succeeds. Focus on stabilizing your primary deck before expanding to additional decks.

1
Day 1: Complete first review session with imported cards. Note any formatting issues.
2
Days 2-3: Identify and repair cards that feel confusing or have unclear prompts.
3
Days 4-5: Run duplicate check and merge cards with identical fronts.
4
Day 6-7: Evaluate your session time. If it increased significantly, investigate card length.
5
By Day 7: Decide whether to continue with Anki, switch fully to Deckbase, or maintain both.

Pro tip: Keep your Anki installation and original decks intact during the first month. This gives you a rollback option if Deckbase doesn't meet your needs.

Need the top-level overview first? Read the published post on migration strategy, then come back to this playbook for execution. Start from Anki import/export guide.