For each registered sentence, automatically analyzes IPA pronunciation, linking, weak forms, and reductions at the word level. Weak forms, C-V linking, flapping, elision, assimilation, coalescence (yod coalescence), dark L, unreleased stops -- 14 sound shift patterns are color-coded and visualized. Tap any pattern to see a detailed explanation of why you couldn't hear it. Adopts the "listen and notice" approach, which phonetics research has proven most effective for non-native learners.

Select any English on a web page, right-click and choose "Hearlixに登録" (Register to Hearlix), and it goes straight into your portfolio. On iPhone, use the Hearlix Shortcut from the share sheet; on Android, use the share menu. Never let a new expression slip away.

Just enter a word or phrase and AI auto-generates meaning, IPA pronunciation, syllable breakdown, 6 example sentences (with sound shift annotations), etymology, synonyms, antonyms, collocations, and CEFR level. Works with any input language. Syntax patterns (e.g., It is ~ for A to do) can be registered with wildcards. Bulk registration is also supported -- import from text files or CSV at once.

Core image, etymology, nuance differences between synonyms, and 6 example sentences with interlinear pronunciation notation (per-word IPA + sound shifts + stress patterns). Popover explanations for each sound shift. TTS audio playback button included.

47+ IPA symbols categorized into vowels, diphthongs, and consonants. Click any symbol to watch a 3D animation video showing mouth and tongue positioning. Voiced/voiceless pair display, R-colored vowels (rhoticity), and pronunciation tips included.

The FSRS-6 algorithm schedules each review for the moment you are about to forget. The Practice Lounge covers meaning, pronunciation, instant composition and dictation. "Today's Review" serves only what is due, "Free Practice" lets you choose, and "Mistakes" keeps what you missed. Meaning and pronunciation graduate after 3 correct answers in a row, and come back as a spot check after 30 days. Every expression is placed on two axes: Memory Strength and Output Power.

AI auto-generates translation challenges for each registered entry. Three difficulty levels (beginner, intermediate, advanced). AI evaluates your answer and provides feedback on areas for improvement.

Pick a theme and how many cards you want, then build a set from your own expressions. Each card shows why it was chosen, based on how much it can still grow in Memory Strength and Output Power. Redraw if you like, confirm the set, and keep the whole board in view during your conversation. Afterwards, record each card as "used while looking" or "used without looking". Using it without looking raises its Output Power.

Enter what you want to say in your native language, and AI prioritizes your under-practiced vocabulary when crafting the English translation. Sentence-by-sentence breakdown, highlighting of used entries, suggestions for noteworthy unregistered expressions, and nuance comparisons of alternative phrasings. Voice input supported. Turn "I wish I had said that" into your next weapon.

Speak into your microphone and Microsoft Azure speech recognition listens. Pronunciation checks score every word and phoneme, and in Instant Composition your spoken sentence becomes your answer. Choose which microphone to use in Settings. The Listening Deck also uses it to check what you repeat aloud, and AI Compose takes voice input in Japanese (for translation) or English (for proofreading). If Azure can't be reached, your browser's built-in speech recognition takes over.

See your streak, today's new entries, completed reviews, and pending reviews at a glance. Cumulative entry trends, mastery distribution, monthly calendar heatmap, and API usage visualized with charts.

All registered vocabulary is converted into semantic vectors using AI (sentence-transformers) and placed in 3D space. Semantically similar words cluster together. Get a bird's-eye view of your arsenal and intuitively spot where the gaps are.