Know exactly where you are
Listening, Speaking and Reading tracked separately
One bar per skill
Most apps give you one level. However, if you are a heritage speaker for example, you might follow a whole conversation and not read a menu. On the other hand, if you can speak and read Mandarin, your reading might already be at the level of newspapers while you can’t speak a word of Cantonese. Each skill has its own level, and every exercise is adapted to that. If you do know Mandarin, just say so in the onboarding steps and we’ll show you common Mandarin to Cantonese mappings.
Watch
Watch what you already watch via YouTube embeds
| Title | Fit | Listening | Pace |
|---|---|---|---|
| Easy Cantonese (HK Interviews University) | Just right | 93% | Natural |
| Easy Cantonese (Macau Interviews on Happiness) | Just right | 94% | Fast |
| 愛回家之開心速遞 03 | Stretch | 88% | Natural |
| McDull 03 | Stretch | 84% | Slow |
| McDull 01 | Hard | 75% | Slow |
Listening = how much of it you already understand by ear.
Paste any YouTube link and we transcribe it, with Jyutping and a gloss over every word. Each video shows how much of it you already understand by ear, so you can sort by fit: Just right to learn from, Stretch to challenge yourself, Hard for later. Tap any word you don’t know to save it.
Speak
Use new words in a real conversation
Your challenge 司機 will steer towards these
Grammar
畀人 + V bei2 jan4
passive: had something done to you
我個銀包畀人偷咗。
Words
去邊度呀?
我去中環呀唔該
中環邊度呀?金融中心、蘭桂坊,定係其他地方?
Talk to a conversation partner about ordering food, getting directions, and more. Turn on a challenge and you get one grammar pattern and two words just past what you can say now. The partner steers toward them and your words stay in the conversation. Words you saved from videos come back here, so what you watch turns into what you say.
How we differ from Duolingo
Our Method: Comprehensible Input
Learn one step at a time. Every new word we recommend is just past where you are (i+1) so you get a little better each time. This is built off linguist Stephen Krashen’s Comprehensible Input theory. Your flashcards are scheduled with Piotr Wozniak’s spaced repetition SM-2 method.
- i+1 Listening/Reading (Input)Watch Fun Videos with Smart Transcripts
- i+1 Vocab (Spaced Repetition)Build Vocab from Videos
- i+1 Speaking (Output)Use Vocab in AI Roleplaying
- i = i+1 ↑ (Growth)Review Newly Retained Content
- Then around again from your slightly higher level
