Complete thoughts, not fragments
The analyst reads the transcript of the whole recording and cuts where a thought starts and ends. Every moment gets a score and a title.
Moment search is free. Tokens pay only for speech recognition.
How the analyst chooses
Three rules instead of “the loudest bit”. A long recording is read in six-minute chunks, so moments are found across the whole recording.
The thought is complete
A clip doesn’t stop mid-sentence: it starts where the thought starts and ends where it ends.
A hook in the first seconds
The title is written from the strongest phrase of the moment and goes at the start of the clip.
Learns from your ratings
Mark a clip “Like” or “Not it” — the analyst remembers which moments suit you.
Ratings affect the next recordings.
The senior model writes more precisely
On the Pro and Studio plans, and with purchased tokens, the senior analyst model works: it finds moments more precisely and writes titles. On Free — the junior one.
Advice about the podcast and listeners0:32:02 — 0:33:05 · 1:03
The most expensive free advice0:32:10 — 0:33:00 · 0:50
Missed? Rebuild it
A rebuild replaces the clip itself: same link, new version, and a new video in the same chat message. Free, except the spoken hook — 7 tokens per 100 characters.
What it looks like on paper
The analyst does what an editor with a highlighter does: reads everything and marks the parts that stand on their own.
0:32:02okay, so what changed in three years?
0:32:10The most expensive advice we got was free:
0:32:18record every week, even when it’s embarrassing.
0:32:33We didn’t listen and lost six months.
0:32:49And when we started, the listeners came by themselves.
0:33:05okay, let’s talk about guests.
Questions about moments
The rest is in the FAQ on the home page.
How many moments will it find?
Depends on the recording. The best by score are built first; the rest wait in “More moments” and can be built in a batch.
Can I pick a moment myself?
Yes, in Editing: the timeline of the whole recording, mark the start and the end — the clip is built with subtitles.
Does moment search cost anything?
No. Tokens pay only for speech recognition: 1 token per minute of recording.
What if I don’t like a clip?
Rate it “Not it” and rebuild: earlier, later, horizontal, without subtitles or with a spoken hook.
Try it on your own recording
150 tokens on first sign-in. No card needed.