What is comprehensible input?

The short answer

Comprehensible input is language you can mostly understand but not completely. You follow the thread. A few words and structures go past you, and the situation, the pictures, the tone of voice and the words around them tell you roughly what they meant.

That gap is where learning happens. If you understand everything, nothing is new. If you understand nothing, there is nothing to attach the new thing to, and the sound goes in as noise. The useful zone is narrow and it moves as you get better.

Where the idea comes from

The term is Stephen Krashen's. In the 1980s he argued that we acquire a language mainly by understanding messages in it, not by studying rules about it, and he wrote the level as i+1: whatever you can already handle (i), plus one step.

Krashen went further than most researchers will follow him. His strong claim was that understandable input is the only thing that causes acquisition, and that grammar study and speaking practice contribute nothing to it. That part is contested and has been for forty years. What is not contested is the weaker version, and it is the one that matters to you: you cannot learn a language you never hear or read, and hours spent understanding it are the hours that move you.

So treat i+1 as a targeting rule, not a doctrine. It tells you how hard your material should be. It does not tell you to stop studying words.

What makes input comprehensible

Difficulty is not one number. Two clips can carry the same vocabulary and be miles apart:

• Speed. A speaker at 110 words per minute leaves gaps between words. At 170 the words run together and the endings disappear.
• Redundancy. A cooking video says "now the onions" while showing you onions. A podcast says it once, in the dark.
• Vocabulary density. The first 1,000 words of a language cover most of everyday speech. A clip that stays inside them is a different task from one about tax law.
• Predictability. You already know how a coffee order goes. That knowledge does half the listening for you.
• Visual support. Faces, hands, objects and subtitles are all context, and context is what turns a word you do not know into a word you can guess.

This is why "watch things at your level" is bad advice on its own. Your level is not a shelf. A B1 learner can follow a slow B2 cooking video and lose a fast A2 argument between friends.

How to tell if something is right for you

The honest test takes two minutes. Play a clip once, without subtitles, and ask what happened.

• You could retell it, and there were a handful of words you did not know: that is the zone. Stay there.
• You got the topic but not the events: slightly too hard. Fine occasionally, tiring as a diet.
• You caught isolated words and no thread: too hard today. Nothing is being learned, and you are practising the feeling of failing.
• You understood every word without effort: too easy to teach you anything new, though it is good for speed and confidence.

A rough rule that has held up: aim for material where you follow about 90 percent on the first pass. Below roughly 70 percent, context stops rescuing you and you are guessing.

What it is not

It is not passive. Sound playing in a room you are not listening to does nothing. You have to be following the meaning. That is the whole mechanism.

It is not "input only". Nothing about comprehensible input says you should never look a word up, never learn a grammar rule, and never speak. Learners who look things up when the guess fails get more out of the same hour. Explicit study is a shortcut into the material, not a competitor to it.

It is not the same as easy. Content made for learners is comprehensible by construction, and a lot of it is comprehensible and useless: scripted, over-articulated, and nothing like what you will meet. Real speech at a manageable level beats fake speech at any level.

It is not a fixed level. The clip that was perfect in March is too easy in June. If your material never gets harder, neither do you.

Where to get it

Two things make comprehensible input hard in practice, and neither is the theory. Finding material at the right level takes longer than watching it. And once you find a creator you can follow, you run out of their videos.

That is what Reelang does. We index real short videos, the ones native speakers post for other native speakers, and grade every clip two ways: the CEFR level of the language in it, and how hard the audio is to follow (speed, connected speech, clarity). Then the feed serves you clips at the edge of what you can handle and moves up as you do.

No quiz, no signup. Pick a language and press play.

If you are still at the very beginning and nothing is comprehensible yet, that is a real problem and it has a real answer: build a small base first, then start watching. Christian Lauer wrote up how to do that in a guest post here, comprehensible input for beginners.

If you would rather hunt for channels yourself, we keep a per-language list of what to search for and which creators are worth your time in how to find comprehensible input on YouTube.