A language learning browser extension adds vocabulary tools to pages you were already going to read: a translation on click, a second subtitle track, a dictionary popup, or a save button. The category is worth understanding as five distinct types, because they solve genuinely different problems and people routinely install three tools that all do the same one. Below is what each type is for, who it suits, and the structural gap they all share.
What a browser extension is structurally good at#
The browser is where most adult learners already meet their target language: news, forums, video, documentation, product pages. An extension's advantage is that it puts a tool exactly there, with no import step and no context switch. That timing matters more than it sounds. A word you look up in the sentence that made you curious arrives with meaning, register, and grammar attached; the same word typed into a flashcard app three hours later arrives naked.
This is also the cheapest possible capture. Selecting a word and pressing one key costs about a second, which is low enough that you will actually do it dozens of times a week. Nothing else in a study workflow has that ratio.
Type 1: click-to-translate readers#
These turn any page into an assisted reading surface. Readlang is the best known: click a word and the translation appears inline above it, drag across a phrase and you get the phrase, and every lookup is stored with its original sentence for later review. LingQ's browser importer works differently, pulling a page into LingQ's own reader where the same click-to-save loop happens inside the app. Generic translation extensions like Mate Translate or Simple Translate cover the lookup half without the vocabulary half.
Who it suits: intermediate readers working through target-language articles. The assumption is that you can already read the page at high coverage and only need the lookup cost removed.
Type 2: subtitle and video tools#
Video is where extensions do something no separate app can, because they operate on the player you are already using. Language Reactor, formerly Learning Languages with Netflix, adds dual subtitles, hover translation, per-line replay, and saved phrases to Netflix and YouTube. Trancy and eJOY do similar work across a wider set of sites. Migaku's extension combines subtitle handling with dictionary lookup and card export.
Who it suits: learners whose listening lags their reading, which is most self-taught learners past B1. If you want a broader look at the options in this specific niche, see Language Reactor alternatives. Be aware that streaming sites change their players often, and subtitle extensions break more frequently than any other type here.
Type 3: word-substitution extensions#
These invert the usual model. Instead of translating target-language text for you, they take a page in a language you already know and replace a fraction of the words with target-language equivalents, marked so you can hover for the original. Toucan is the widely used example, and a handful of smaller tools do the same thing.
The mechanism is the diglot weave, described by anthropologist Robbins Burling in 1968: embed unfamiliar words in sentences the reader already understands, so context carries the meaning. It is genuinely effective for recognition and demands nothing of you beyond browsing. Its limits are structural rather than product flaws: the substitution density has to stay low to keep pages readable, and because the surrounding sentence is in your own language you get vocabulary without syntax. Readlang vs Toucan compares this philosophy against active lookup in detail.
Type 4: popup dictionaries#
The most technically precise category, and the one power users care most about. Yomitan, the maintained successor to Yomichan, shows a dictionary entry when you hover over a word, handles Japanese conjugation and deinflection properly, and can push a card into Anki through the AnkiConnect add-on. 10ten Japanese Reader does the hover-dictionary job for Japanese, and Zhongwen does it for Chinese with pinyin and tone information.
Who it suits: learners of languages where segmentation or script is the barrier. For Japanese and Chinese in particular, a good popup dictionary is not a convenience, it is the difference between a page being readable and being an image. These tools are free, open source, and configurable, but they expect you to set up dictionaries yourself.
Type 5: save-to-app extensions#
The last category does one job: get the word out of the browser and into wherever you actually review. Yomitan plus AnkiConnect is the classic pipeline. LingQ's importer serves the same purpose for its own library. LingoBlend's Chrome extension adds a right-click "Save to LingoBlend" on any selected text, which translates the word, detects its base form, and drops it into your dictionary, where it is scheduled for the app's review games and appears on your phone.
Who it suits: anyone who already has a review system and is losing words between the page and the queue. A fuller walkthrough of the capture-to-review loop is in saving vocabulary from any website.
The five types compared#
| Type | What it does | Examples | Best for | Built-in review |
|---|---|---|---|---|
| Click-to-translate reader | Inline translation on click, saved with the sentence | Readlang, LingQ importer | Intermediate readers | Yes, in-product |
| Subtitle and video tool | Dual subtitles, hover translation, replay | Language Reactor, Trancy, eJOY, Migaku | Listening practice | Partial, saved lines |
| Word substitution | Swaps native-language words for target ones | Toucan | Beginners, ambient exposure | Practice on saved words |
| Popup dictionary | Hover definition with deinflection | Yomitan, 10ten, Zhongwen | Japanese, Chinese readers | No, exports to Anki |
| Save-to-app | Sends the word to your review system | LingoBlend, AnkiConnect pipelines | Anyone with an existing queue | In the receiving app |
What browser extensions are structurally bad at#
Three limits apply to the whole category. First, scheduling: a browser tab has no idea that a word you looked up on Tuesday needs to reappear on Friday. Spacing is the single best-evidenced technique in vocabulary learning, from Ebbinghaus in 1885 through modern meta-analyses of distributed practice, and almost no extension implements it. Highlighting a word feels like learning it, which makes this gap easy to miss; spaced repetition for language learning covers why the review step is not optional.
Second, mobile. Chrome on Android supports no extensions at all, while Safari on iOS and Firefox on Android support only a subset. Since most reading time now happens on a phone, a desktop-only capture tool covers a minority of your input unless it syncs somewhere.
Third, fragility and permissions. These tools must read page content to work, so they request broad access, and they break when sites change. Chrome's migration to Manifest V3 forced every extension to be rewritten, and some never were. Install from official stores, check what a tool does with page data, and expect occasional breakage.
A stack that closes the gap#
The practical arrangement is one capture tool plus one review system, chosen so the handoff is automatic. A popup dictionary or click-to-translate reader handles the moment of encounter, and everything it captures lands somewhere that schedules it. What you want to avoid is five extensions that all highlight words and none that ever ask you to recall one. If your reading is happening on a phone as often as a laptop, choose a capture tool whose destination is an app you already carry, so the words meet you again without a second import step.