Comprehensible Input
— Krashen, 1985You acquire a language by understanding content slightly above your level. Bilingual sentence pairs make any video i+1.
AUDIO-FIRST LANGUAGE APP
Listen to anything. Twice.
Any YouTube video. Any text. Each sentence plays in the original — then once in your language. Headphones on, twenty minutes a day, the language gets in.
EN
This program is brought to you by Stanford University.
zh
本节目由斯坦福大学为您呈现。
Demo · Steve Jobs' 2005 Stanford Commencement Address
AI-narrated preview. Paste a URL below to hear the original.
Free signup · No credit card · Robot voice free
— OR BRING YOUR OWN
Two ways in. One way to listen.
Drop a YouTube URL or paste any text. Steep lines it up sentence by sentence, plays each one in both languages, and saves it to your library.
Robot voice is free. No credit card required. Signup takes 30 seconds.
— WHY IT WORKS
Four findings. One method.
Steep stitches together what second-language-acquisition researchers have been writing for forty years. None of it is new. The integration is.
Listening-Reading
— Atamian; ArguellesWhen the original and its translation arrive in the same beat, your brain maps sound to meaning. No conscious translating.
Shadowing
— Arguelles; ATA standardRepeat any sentence aloud in the gap after it plays. The interpreter's training method, in your ear.
Affective Filter
— Krashen, 1982Anxiety blocks acquisition. So no quizzes, no streaks-as-shame, no forced output. You just listen.
— A SESSION
Drop. Listen. Walk away with an hour of input.
One session is twenty minutes. You don't sit down to study. You put your headphones on and start the day.
01
Drop in what you want to hear.
A YouTube URL or paragraphs of text — anything in the language you're learning.
02
We line every sentence up.
Steep extracts and aligns each sentence with its translation, sentence by sentence.
03
You listen, in two languages.
Original audio, then your language, then the next sentence. No pausing. No translating in your head.
04
Loop, shadow, or just walk.
Tap a line to loop it. Repeat aloud if you want to. Stop when you want. Sessions save to your library.
WHY MOST APPS DON'T WORK
Two years of tapping an app. You still can't follow a YouTube video.
That isn't a discipline failure — tapping isn't listening.
Steep is the listening.
— PRICING
Two tiers. That's it.
Robot voice, free. Human voice, in Pro.
Free
$0/mo
Robot voice
3,000 credits / week (≈ 1,600 words of Standard voice)
- · Sentence Mode on any video or text
- · 30-day history
- · Save sessions
No credit card required
Start freePro
$7.50/mo, billed $89.99/yr
Human voice (Google + ElevenLabs)
60,000 credits / month (≈ 32,000 words Standard or ≈ 4,800 words Realistic)
- · Up to 5,000 chars per request
- · Unlimited history
- · Audio downloads
Save ~42%. Cancel anytime
Choose Pro— START YOUR FIRST SESSION
Twenty minutes. Headphones on.
Free to start, no credit card. Bring a YouTube video you've wanted to watch, or any text you've been meaning to read.