Comparison
Wanilog and jpdb: stats for the WaniKani half of your journey
jpdb and Wanilog answer different questions. jpdb asks "what vocabulary did I just see in this anime, this visual novel, this web novel - and when should I review it?" Wanilog asks "where am I on WaniKani's curriculum, when will I finish, and what can I read along the way?" If you live in immersion content, jpdb is built for that life. If you are working through the WaniKani 60-level path, Wanilog gives you the stats around it. Many learners use both.
What jpdb does well
- It treats immersion as the syllabus. Paste a script, a VN, or a web novel and jpdb extracts every word, threads it into a deck, and schedules it. The content drives the deck, not the other way around.
- A library of 21,000+ prebuilt decks. Anime, novels, visual novels, web novels. Chances are someone has already mined whatever you are reading.
- A modern SRS algorithm. jpdb moved off SM-2 to a machine-learning-derived schedule that handles lapses and irregular reviews more gracefully.
- i+1 sentence cards on autopilot. Cards arrive in sentences where every other word is already known. Hard to engineer at scale; jpdb does it.
- A 130M+ example-sentence corpus. Real Japanese, not invented practice text.
Side-by-side feature matrix
| Feature | Wanilog | jpdb |
|---|---|---|
| Setup and hosting | ||
| Hosted, no install | Yes | Yes |
| Free tier with full feature access | Yes | Partial |
| Demo mode (no key, no signup) | Yes | No |
| Self-host option | No | No |
| Stats and analytics | ||
| WaniKani SRS stage distribution | Yes | No |
| Per-WK-level accuracy chart | Yes | No |
| WK leech list with severity | Yes | No |
| Leech trainer (multiple algorithms) | Yes | No |
| WK burn velocity (week / month) | Yes | No |
| Workload forecast (lessons-per-day slider) | Yes | No |
| Peak review day surfaced | Yes | No |
| Coverage and forecasting | ||
| Level 60 finish-date projection (WaniKani) | Yes | No |
| Statistical projection with p25–p75 band | Yes | No |
| Goal level + goal date with on-track banner | Yes | No |
| JLPT N5–N1 kanji coverage | Yes | Partial |
| Joyo coverage (grades 1–9) | Yes | No |
| Top-2500 frequency-list coverage | Yes | Partial |
| Reading coverage on live NHK headlines | Yes | No |
| "Can I read this?" arbitrary text tool | Partial | Yes |
| Vocabulary mined from anime / novels / VNs | No | Yes |
| Prebuilt decks from immersion content | No | Yes |
| Per-novel coverage on uploaded text | No | Yes |
| i+1 sentence cards from a 130M+ corpus | No | Yes |
| UX and device support | ||
| Installable PWA | Yes | Partial |
| Offline shell | Yes | No |
| Phone-first layout | Yes | Partial |
| Dark mode | Yes | Yes |
| Color palettes (4) | Yes | No |
| Cinematic level 1-to-current replay | Yes | No |
| Share-card export (PNG) | Yes | No |
Wanilog and jpdb solve different problems. They overlap on a few coverage tools and diverge everywhere else. If you are on WaniKani, Wanilog adds the stats jpdb does not aim to provide. If you immerse in real content, jpdb adds the mining and scheduling Wanilog does not aim to provide. Most serious learners use both.
Where they diverge
The source of vocabulary
jpdb's deck is whatever you put into it - the script of the anime you started last night, a VN, a Wikipedia article. Vocabulary grows from your media diet. Wanilog has no opinion on what you read; it surfaces only the kanji and vocabulary inside WaniKani's 60-level corpus, and reports on how far through that corpus you are.
The SRS itself
jpdb owns the schedule - it is your reviewer, interval-picker, and scoring system. Wanilog owns no schedule. WaniKani's SRS is the schedule; Wanilog reads it from the API and visualises it. So "which SRS is better" is not a question this page can answer - Wanilog is not an SRS at all.
Reading coverage
Both tools care about whether you can read things, but they answer different versions of the question. jpdb says "of the 8,932 unique words in this novel, you know 78%." Wanilog says "of the kanji in today's NHK Easy headlines, you have taken 64% to Guru." The jpdb answer is sharper for a specific book; the Wanilog answer is sharper for a daily moving target tied to your WaniKani progression.
Finish-date projection
jpdb does not project a finish date because there is no finish line - the deck grows as you read more. WaniKani's 60-level structure has one, and projecting it is something Wanilog was built around: a p25-p75 confidence band fit to your actual level history, with resets and breaks filtered out, goal-level aware. Try the no-key calculator version.
Account and data model
jpdb is a server-backed account - your decks, your reviews, and your progress live on their infrastructure. That is the right shape for a tool that has to remember every word you ever mined. Wanilog is fully client-side - no account, no backend, your WaniKani API key sits in your browser's localStorage, and the dashboard rebuilds itself from the WaniKani API on every device.
When jpdb is the better fit
- You are immersion-first, not curriculum-first. If your daily Japanese habit is "watch a thing, read a thing, mine the words," jpdb's whole workflow is built around that.
- You are not on WaniKani. Wanilog reads exclusively from the WaniKani API. Without a key, it has nothing to draw on.
- You want a parser, not a dashboard. jpdb tokenises, conjugates, and threads sentences. Wanilog scores; jpdb explains.
FAQ
Can I use Wanilog and jpdb together?
Yes, and many WaniKani learners do. jpdb handles vocabulary mined from your immersion content; Wanilog tracks your WaniKani progression. They do not overlap on the SRS itself, so there is no double-review problem.
Does Wanilog import vocabulary from outside WaniKani?
No. Wanilog reads only from the WaniKani API, so it only knows about WaniKani vocabulary. If you want vocabulary mined from anime, novels, or visual novels, that is what jpdb is for.
Why is Wanilog’s reading coverage different from jpdb’s?
They measure different things. jpdb tells you the percentage of unique words you know in a specific text, anchored to vocabulary. Wanilog tells you the percentage of kanji you have taken to Guru in live NHK Easy or NHK News headlines, anchored to your WaniKani kanji set.
Does Wanilog support immersion tracking?
No, and it is not planned. Immersion mining is jpdb’s domain and they do it well. Wanilog focuses on giving WaniKani users better stats on the curriculum they are already paying for.
Is Wanilog free?
Yes. No paid tier, no ads, no account. Wanilog works with free WaniKani accounts (capped at level 3) and paid accounts (capped at level 60).
Does Wanilog work offline?
Yes. Wanilog is an installable PWA with a Service Worker. After your first sync, the dashboard, coverage pages, and item drill-downs all work without a connection. New WaniKani reviews still need the API.
Will my WaniKani progress sync to jpdb, or vice versa?
No. The two services do not talk to each other. You review on WaniKani and on jpdb separately, and each tool reports on its own world.
Can Wanilog parse a novel I paste into it?
Wanilog’s "Can I read this?" tool scores arbitrary Japanese text against your WaniKani kanji set, which gives you a fast readability percentage. It is not a full tokeniser - jpdb’s parser is more thorough when you want a per-word breakdown.
Where does my WaniKani API key live?
In your browser’s localStorage, never on a Wanilog server, because Wanilog has no server beyond a tiny rate-limited NHK news proxy. The key only travels between your browser and api.wanikani.com.
Already on WaniKani? See what your stats look like.
Paste your API key, or open the demo first. Both take under a minute, and nothing leaves your browser.