Materials from your own audio
Load an mp3 / m4a, pick the region on the waveform, and AI transcribes it into a script with per-word timings.
SHADOWING × AI COACH
From hearing it
to saying it.
Turn your own audio into shadowing materials, record yourself, and get word-by-word AI feedback on pronunciation, rhythm and intonation — plus the one thing to fix next.
- Preset materials in many accents
- 103 segments × 3 speeds
- AI scores pronunciation, rhythm and intonation
- Word by word
- Pick a region, let AI transcribe it, practise
- Your own audio
FEATURES
Everything shadowing needs, in one place.
From building materials to recording, AI analysis and review — the whole shadowing loop on one screen.
Karaoke highlight and script toggle
The current word lights up as the audio plays. Once it feels familiar, hide the script and shadow by ear.
Record and let AI analyse
Your voice is recorded alongside the model and scored word by word on pronunciation, rhythm and intonation. The AI coach names the one thing to fix next.
Presets in many accents
Indian, Singaporean, Korean, Japanese, Chinese, American, British and Australian speakers and more, each at 120 / 140 / 160 WPM. Dialogues can be practised as role-play.
Vocabulary book
Save any word from the script with one click. AI adds a meaning that fits the context and an example sentence.
Learning history
Score trends and practice time are recorded automatically. The more you practise, the clearer the progress.
AI FEEDBACK
A coach that names the one thing to fix.
Your take is aligned to the model word by word and comes back as an overall score, a three-way breakdown and colour-coded words. Instead of a long critique, the coach gives you a single focus for the next take.
- Pronunciation — how closely each recognised word matches the model
- Rhythm — per-word timing and pauses
- Intonation — pitch movement compared with the model
HOW IT WORKS
Three steps, starting today.
- STEP 01
Prepare material
Pick a region of your own audio, or add a preset to your library.
- STEP 02
Shadow and record
Speak along with the model. Slow it down and loop it as often as you like.
- STEP 03
Improve with AI analysis
Read the per-word evaluation and the coach's comment, then record the next take. Your scores stay in the history.
PRESETS
Tune your ear to the world's English.
Every preset comes at 120 / 140 / 160 WPM: start slow, then move up to real-world speed. Dialogues let you pick a role and answer the other speaker's lines.
- American
- British
- Australian
- Indian
- Singaporean
- Korean
- Japanese
- Chinese
- Canadian
- Irish
Chai Before the Meeting
DialogueBeginner
120 WPM140 WPM160 WPM
0:27 · 54 words · 120 WPM
Speakers: Priya · Indian / Arjun · Indian
So one cup of chai before the client call? — One is never enough. Not on a Monday.
Lunch at the Hawker Centre
MonologueIntermediate
120 WPM140 WPM160 WPM
0:24 · 56 words · 140 WPM
Speakers: Wei Ling · Singaporean
The queue tells you everything. If the auntie's stall has a line at eleven, that's where you eat.
Stand-up at a Tokyo Startup
DialogueAdvanced
120 WPM140 WPM160 WPM
0:22 · 58 words · 160 WPM
Speakers: Ken · Japanese / Emma · British
Quick one — the release slipped to Thursday. — Fine, but then QA needs the build by tomorrow noon.
Filter by type, level, speed and accent. Switch the speed on the card and add it to your library in one click.
PRICING
14-day free trial, then the Pro plan
Every feature is free for 14 days after you sign up. Move to Pro when you want to keep going.
Pro14 days free
¥980 / month (tax incl.) · ¥9,800 / year
The exact price is shown at checkout.
What's included
- Materials from your own audio (AI transcription)
- AI analysis of your takes — pronunciation, rhythm, intonation and coach comments
- Preset materials in many accents (3 speeds, role-play)
- Vocabulary book, Japanese translation and learning history
- Cancel anytime
FAQ
Frequently asked questions
Which audio files are supported?
mp3 and m4a (wav works too). One material can cover a region of up to 10 minutes. Many people export the audio of a podcast or a video.
Can other people hear my recordings?
No. Your audio and recordings are stored in your account only, and row-level security (RLS) in the database keeps everyone else out. You can delete materials and recordings at any time.
Which languages are supported?
Materials are in English. The interface can be switched between Japanese and English.
Is there an iPhone app?
It is in preparation and will share your account with the web version, so materials, recordings and history stay in sync. For now, the web app also works in your phone's browser.
Can I cancel at any time?
Yes. Cancel whenever you like; you will not be charged after the current period ends.
Start with 14 days, free.
All you need is an email address.