A Klap alternative where the list comes before the clips
Both read a long recording and both propose shorts. The difference is what you get handed: a batch of finished clips with a number on each, or every candidate with its four scores, its timecodes and the sentences it was cut from — before anything is rendered.
What we are not claiming
Not that it picks better moments. Nobody can demonstrate that to you on a landing page, and any product claiming it is asking you to take the one thing that matters on trust. Selection quality is a thing you find out on your own footage, in an afternoon, and both of these have a free way to let you.
What is checkable is what you are shown, and when. Here, the whole candidate list arrives before anything is cut: four scores broken out, the in and out timecodes, and the sentences each candidate was cut from. You keep or you pass, and the renderer never runs on a passage you did not approve.
That changes what you can do when it is wrong. A score with no reasoning behind it can only be accepted or ignored. A score with its four pillars and its source sentences can be argued with — you can see that a candidate scored high on hook and thin on value and decide that is not the clip you want, before spending anything on making it.
How a moment is chosenWhere the two differ
Sentences rather than ticks and dashes. This is the closest product to ours in intent, so the table sticks to what a user is shown rather than to how either one works underneath.
| What differs | InTheClips | Klap |
|---|---|---|
| What arrives first | A ranked candidate list. Four scores, timecodes, transcript and one line of reasoning per candidate; rendering is a second, separate decision. | Clips, produced from the recording and reviewed as video. An announcement on 13 August 2026 described work on how the selection is made; what a user is shown is still the output. |
| What you can inspect | The reasoning. The pillars are stored on each moment rather than inferred for display, and the passage is on screen next to them. | The result, and the score attached to it. |
| Ways in | A YouTube link or a file. A watch page, a Short, a live or an embed address is read without a copy of the video being kept — audio for the transcription, then only the seconds of the window each render needs; links are capped by the day, uploads are not. | Links from the sources named on its own site — read that list there rather than here, and expect it to be wider than the one platform we take. |
| Building on top of it | No public API. What leaves the product is an MP4 you download. | A public API, which is a real advantage for anyone wanting this inside their own pipeline. |
| What is metered | One credit balance, not two meters: minutes of source analysed and clips exported past the included batch — 80 on Starter, 220 on Creator, 540 on Studio — both spend the same number, at one credit a minute and three a clip. | Its own units, on its own site — a number copied here would be stale by the time you read it. |
Their column is from Klap’s own public description of the product, read on ; ours is from the code that runs this one.
The parts of this you can check yourself in one session
Every one of these is visible in the product on a trial, which is the only reason they are worth listing.
The back half is ranked too
Thirty-minute windows scored in parallel with three minutes of overlap, then one judge ranking the pooled finalists — because a single pass anchors on what it read first and stops picking from the end of a two-hour file.
A boundary never lands inside a word
The model quotes the opening and closing words verbatim and the times are read off those exact words. Anywhere is allowed except mid-word.
A cold reader checks the cut
A second pass sees only the clip’s own lines and answers as somebody who just landed on it. If it cannot follow, the cut is made again — twice at most, then the clip is kept with its score capped and a line saying what is missing.
Up to three pieces, in order
A moment can be assembled from the setup a minute earlier plus the moment itself, with the detour between them dropped. Drag the handles by hand and that assembly is discarded, because it was reasoned about a window that no longer exists.
This is the competitor closest to us in intent, so the list of what they do better is the one worth reading twice.
A link here is youtube.com or youtu.be and nothing else: a Twitch VOD, a Vimeo file, a Zoom recording or a Drive folder is refused, and the way in for those is an export and an upload. Klap names its own list of sources on its own site and it is worth reading, because breadth of intake is the part of this we have not matched. What we do take is also counted by the day — three on the trial, a hundred at the top — where an upload never is.
They publish an API. There is no public API here, so anything that needs clipping to happen automatically inside somebody else’s system is a case this cannot serve at all.
The candidate list is ten minutes of your attention on a two-hour episode. If the point of buying an auto-clipper was to not spend that, then the thing this product is built around is the thing you are trying to avoid, and that is a fair position.
A tool that produces a batch and moves on is arranged for output. Ours is arranged for a decision per candidate, and the units are priced accordingly — read both against how many hours a month you actually put through.
Questions
The ones worth answering before an afternoon is spent finding out.
Which one picks better moments?
We will not answer that, because we cannot prove it and neither can anyone else in this category. What we would say is that the question is answerable by you in one afternoon: put the same recording through both free tiers and compare the passages that come back. That test is worth more than every comparison page on the internet, this one included.
Can I paste a YouTube link?
Yes — a watch page, a Short, a live URL or an embed. No copy of the video is kept here: the transcription reads an audio-only stream and each render pulls only the seconds of its own window, so a long recording is never downloaded in order to be uploaded again. What that trades is worth knowing — a link goes on depending on the video staying at that address, and a file you send does not. Two conditions come with it. You confirm you hold the rights before the import starts, and the sentence you were shown, the time and the address are stored with the video. And imports are counted by the day — three on the trial, twenty on Starter, fifty on Creator, a hundred on Studio — while uploads are not counted at all. Links are YouTube only: for a Twitch, Vimeo or Zoom recording the way in is still the file.
Is there an API?
Not a public one. Everything happens in the product and what leaves it is an MP4 you download. If clipping has to run inside your own pipeline, that is a case this does not serve today.
What does the finished clip look like?
Boxed 1:1 with the words composited behind the speaker, who is matte-cut and sits over the text, or full-frame 9:16 with word-by-word captions timed off the transcript. In both the crop follows whoever is speaking, and a banner guard punches in only when it measures a lower third already baked into the footage.
Is there a free tier to try it on?
Thirty minutes of video, once, no card — one half-hour recording and every clip that comes out of it. It does not refresh monthly on purpose: a recurring free allowance is a recurring bill, and shrinking one quietly later is worse than being straight about it now.
What should I check before subscribing to either?
Whether a renewal is announced before the charge, whether cancelling is a button rather than an email, and what the billing unit actually counts. Those three are behind most of the complaints written about tools in this category — most of them, not one — and they are all answerable before you pay. Ours cancels from the account page without emailing anybody. On the reminder we are not yet where we would like to be, and saying so on a page that tells you to check is cheaper than being caught at it.
Coming from something else
More of these, written the same way: what changes, and what you would be giving up.
