Immersion trackers compared: what each one actually counts

Every immersion tracker ends in the same place: a number of hours. The differences that matter are not the charts or the streaks but how that number is made, because the method decides what the number contains. There are four ways to make it. You type it in afterwards. You press a timer before and after. A tool watches which website is open. Or a tool listens to what is playing and counts only the language you are learning. Each is honest about something and blind to something else, and the right tracker is the one whose blind spot you can live with.

The short answer. If you track reading, speaking, tutoring and flashcards as well as listening, you need a manual log, and LinguaTrackr, TrackMyLang and LangTrack are all free, well made and differ mainly in feel. If you already use Toggl for work, its timer and tags will do. If nearly all of your input is video in a browser, a manual log will fail you within a month, not because the log is bad but because it depends on remembering to fill it in, and a tracker that reads time from the player is the one that keeps counting when you forget. Kiku Hours is the one built for that case: it counts playback rather than minutes a tab was open, identifies the language of each video by listening on your own computer, and holds out ads, pauses and videos in a language you already speak. It counts nothing you did away from the browser.

This comparison was checked against each product’s own site on 19 September 2026. Products change; if something here is out of date, the support page is where to say so.

The four ways of making the number

Typing it in. A spreadsheet, or an app with an “add session” form. The number is whatever you remember, rounded to what feels right. It is perfectly accurate on the evening you fill it in and drifts from there; a four-day gap back-filled from memory is a guess wearing a record’s clothes. Its great virtue is that it can hold anything: a tutoring session, a chapter of a novel, an hour of Anki. Nothing automatic can do that.

A timer. Press start, press stop. More honest than memory, because the clock does not round up, and still blind to what happened between the presses: the ad break, the ten minutes you spent looking up a word, the tab you left open through dinner. Timers also depend on the second press, and everyone who has used one has found a timer still running the next morning.

Watching a domain. A browser extension that notices youtube.com is open and counts. This is where automatic tracking usually starts, and it is the reason “minutes watched” misleads: an open tab is not a playing video, a playing video is not necessarily in your target language, and the ad in the middle was in neither.

Listening. A tool that reads time from the video player’s own clock, so only playback counts, and identifies the language of each video from its audio, so only your target language counts. It is the only method that produces input hours rather than watch time without asking you anything. Its blind spot is everything that is not video in a browser.

The trackers, one by one

A spreadsheet. Free, yours forever, and it can hold any column you think of. Everything in the case against spreadsheets applies: it depends on you, every day, after the video has ended. Most people who start one stop within a month. The ones who keep it up are the ones who would have counted anyway.

Toggl Track. A general-purpose time tracker, not a language tool, with “manual & real-time timers” on a free plan that allows an unlimited number of users; the paid plans, from $9 per user a month, add team reporting you will never need. Immersion learners use it with tags, one for active and one for passive listening, and read the percentages off the reports. If you already time your work in Toggl it is the obvious choice. It knows nothing about languages and counts whatever is between the presses.

LinguaTrackr. A web app for logging sessions by hand, with activity types that cover the whole of language learning: listening, reading, speaking, writing, vocabulary, grammar, immersion and apps. Ninety-eight languages plus any you add yourself, optional metrics such as pages and episodes, and an export of your full history as CSV or JSON. Free, with no premium tier mentioned and no credit card asked for. It is the fullest manual log here, and it is manual: it records what you tell it.

TrackMyLang. A web app with a live timer as well as manual entry, eight categories and more than eighty subcategories, an Anki integration and a watch list. Its distinctive choice is being local-first: sessions are “stored in your browser” with no account, and cloud sync is optional. The free tier has unlimited tracking; Pro is $3 a month for analytics and automatic sync, and adds a URL auto-fill that pre-fills a log entry’s title and duration from a pasted YouTube link. Note what that duration is: the length of the video, not the time you listened to it.

LangTrack. The simplest of the manual logs, and proud of it: “Hit the timer when you start studying, or add minutes manually after. Either works.” Tags for reading, listening, speaking and watching, weekly bars, a yearly heatmap, and every feature free with no limit on languages or sessions. If the reason your spreadsheet died was friction, this removes most of it. It still needs you to press the button.

Track your Input. A free, open-source Firefox add-on that “automatically tracks your YouTube watch time or any other videos on other websites for the language you are learning”. It is the closest thing here to Kiku Hours in intent. The listing names fifteen languages and does not say how a video is assigned one; it asks for access to your data on all websites; and on the day of checking it had one user, no reviews and a last update from October 2025. It counts watch time, which is the domain-watching method with its blind spots.

Polylogger belongs in this list, being the log many Japanese learners use, but its site would not let us read it on the day of writing, so it is left out rather than described from memory.

Kiku Hours. A free extension for Chrome, Firefox and the other Chromium browsers that reads time from the player, so ads, pauses and seeks add nothing and a hidden tab still counts; identifies the language of each video by listening, on your computer, among 106 languages; and holds out time in the languages you say you already speak, showing the held-back amount beside the total rather than deleting it. When a video’s language cannot be identified its time is not counted, because a padded total is worse than a short one. Everything stays in your browser with no account, and can be exported or backed up. It works on YouTube, Twitch, Kick, Vimeo, Dailymotion, TED, Bilibili, Niconico and Netflix, and nowhere else: a podcast app, a paperback or a conversation is invisible to it, and it has no form for adding them by hand yet.

Side by side

Tracker	What it is	How the number is made	Knows the language?	Ads and pauses	Price
Spreadsheet	A file	You type it	If you add a column	If you remember	Free
Toggl Track	Time tracker	Timer or manual	No; tags	Counted	Free; teams from $9
LinguaTrackr	Web app	Manual sessions	You choose	Not applicable	Free
TrackMyLang	Web app	Timer or manual	You choose	Counted by the timer	Free; Pro $3 a month
LangTrack	Web app	Timer or manual	You choose	Counted by the timer	Free
Track your Input	Firefox add-on	Watches video sites	Not stated	Counted	Free
Kiku Hours	Browser extension	Reads the player’s clock	Identified by listening	Excluded	Free

“Counted” means the tool has no way to know the difference; it is not a criticism of a timer for being a timer.

What the manual logs do that Kiku Hours cannot

It would be dishonest to end without this. A manual log covers your whole learning life: the tutor, the novel, the flashcards, the podcast on a walk, the film at a cinema. Kiku Hours counts video in a browser and nothing else, and until it has a way to add other time by hand, a learner who reads as much as they listen will get a fuller picture from LinguaTrackr or TrackMyLang. Toggl is right for anyone who already lives in it. LangTrack is right for anyone who wants the least possible between them and a number. The trade is exact: the manual logs will record anything and depend on you; the listening tracker depends on nothing and records only what it can hear.

Which to pick

If most of your hours are YouTube, Twitch or Netflix in a browser, use a tracker that reads the player and listens, and let it run. If most of your hours are elsewhere, use a manual log, and pick the one whose form you will actually fill in. If you are somewhere between, do both for a month and see which number you still trust at the end of it. The one you trust is the one that will still be running next year, and a tracker that is still running is the only kind worth having.