What your video’s words reveal
Every speaker has familiar words. Yap Lab helps you notice them by counting the captions in a YouTube video.
Start with a video link
Paste a watch, Shorts, live-video, or youtu.be link on the home page. A public video with accessible captions works best. Yap Lab fetches existing captions through Supadata and creates a report with word bubbles, a leaderboard, repeated phrases and possible fillers.
Why a small word can dominate
Words such as “the,” “and,” and “you” connect sentences. They occur often in almost every English conversation, so the main bubbles filter common words out. Open “A little more word nerdery” and turn on “Include common words” to see the full leaderboard. This filter mainly covers English; reports in other languages may need more interpretation.
How counting works
Capital letters become lowercase, punctuation separates words, and curly apostrophes are normalized. “Hello” and “hello” therefore count together, while “create” and “creating” remain different words. Pure numbers are excluded. The bubble sizes reflect relative frequency within that video, so a large bubble does not mean the speaker uses that word more than everybody else.
Explore the context
Tap a word to see caption snippets containing it. The timestamps come from the captions. Selecting one opens the embedded YouTube player near that moment. Captions group words into segments, so a timestamp marks the start of the segment, not necessarily the exact word.
Repeated phrases and possible fillers
Yap Lab looks for recurring sequences of two to four words and shows a few distinctive examples. A phrase can cross a caption boundary. The filler list highlights words such as “um,” “like,” and “basically”; context decides whether they are actually fillers. A caption service may omit hesitations, and a word like “like” can carry meaning. Treat these counts as clues, not a speaking score.
Keep and share the result
Once an analysis is saved, its report link opens the same caption snapshot without fetching it again. You can share that link or generate a square PNG report. Anyone with the report link can view it. A saved report may differ from captions later edited on YouTube.
Free usage and unavailable videos
Each network can try 3 new videos per day, resetting at midnight UTC. Opening saved reports does not use that allowance. Failed new-video attempts also count, because the caption service may still process them. A separate site allowance keeps overall costs bounded.
Private, deleted, restricted or captionless videos may not work. A report reflects the available caption track and is not a verified transcript or professional assessment. Yap Lab does not generate replacement audio transcripts.