Make Autocut
Try free
By Yohann Kipfer · Last updated July 2026

A Descript alternative for the one job you probably opened it for

Descript is a full editing suite behind a monthly subscription and a desktop install. Most people open it to do one thing: cut the silences and tighten a talking-head take. That doesn’t need a suite. Make Autocut does exactly that job in a browser tab — no install, no account for the first file, and the upload is deleted an hour later.

Straight version: if you edit by editing a transcript, or you need “um” and filler words gone, keep Descript — we genuinely don’t do that. If you want the dead air cut, cheaply and in the browser, and the edit handed back to you, that’s this. Free tier: 5 min, 200 MB, one file a day, no card.

Why people go looking for a Descript alternative

Descript is good software — that’s not the reason people search for something else. The reasons are almost always about the shape of it, not the quality.

It's a subscription for something you do occasionally

A monthly plan makes sense if you live in the editor. If you cut a few files a month, you're renting a whole suite by the month to use one feature — silence removal — that you could do without a plan at all.

It's a desktop install

Downloading and installing an editor is a non-starter on a locked-down work laptop, a Chromebook, or someone else's machine. A browser tab has no such problem.

The learning curve is more than the task

Transcript editing, scenes, layers, AI features — it's a lot to learn when the thing you actually wanted was “take the pauses out.” The tool-to-task ratio is off.

The file lives in someone's cloud

For a sensitive interview, a project sitting in a cloud library until you remember to delete it is more exposure than you want. Upload, cut, auto-delete is the cleaner shape.

Descript vs Make Autocut, line by line

This isn’t a table where every row has our checkmark. Descript wins the rows about scope — it does far more than we do, on purpose. We win the rows about cost, access and privacy for the one job we’re built for. Read both columns.

Descript compared with Make Autocut across pricing, install, languages, privacy and scope
 DescriptMake Autocut
What it's built forA whole editing suite: transcript-based video/podcast editing, AI voices, screen recordingOne job: detect and remove silence, hand back the cut
Pricing modelSubscription — free tier, then roughly $16 / $24 / $40 a month by transcription hoursFree tier (no card), then Pro; no per-hour transcription meter
Where it runsDesktop app you download and install (macOS / Windows), plus a web editor100% in the browser — nothing to install, works on a locked-down work laptop
Account needed to startSign-up required before you touch a fileFirst file needs no account at all
Learning curveReal — it's a transcript editor, a timeline, scenes, layers, AI features to learnUpload, look at the waveform, click once
LanguagesTranscription in ~20+ languages; quality varies by languageEvery language — detection is volume-based, no speech model to be trained or not
Filler-word / 'um' removalYes — removes “um”, “uh”, filler and repeated words from the transcriptNo — an “um” is sound, a loudness detector keeps it
What you get outRendered video/audio, published pages, and its own exportsCut file (MP4/MP3) or an editable NLE timeline (Premiere, FCPXML, EDL, OTIO)
Your uploaded fileStored in your cloud project until you delete itDeleted automatically about an hour after processing
Watermark on freeSome free-tier exports carry a watermarkNo watermark, ever — including the free tier

Descript’s prices and tiers are what it published around July 2026 and are starting points, not quotes — check descript.com, they move faster than this page. The rows that stay true regardless of pricing are “where it runs”, “filler-word removal” and “your uploaded file”: a desktop suite that edits transcripts and keeps a cloud project, versus a browser tool that cuts silence and forgets your file.

When Descript is the right tool — don’t switch

A comparison written by the challenger always hides the boxes it can’t tick. Here are ours, out loud. If one of these is your real problem, close this tab and keep Descript.

You edit by editing the words

Descript's whole idea is a transcript you edit like a document — delete a sentence in the text, the video cuts with it. If that workflow is why you're here, nothing on this page replaces it. We have no transcript at all; detection is pure loudness.

You need filler words and stutters gone

“Um”, “uh”, “like”, “you know”, repeated false starts — Descript finds those in the transcript and strips them. A silence cutter can't: those are sound above the noise floor, not gaps. If your recording is clean but verbally messy, Descript is the right tool and we're honestly not.

Screen recording, AI voices, overdub, collaboration

Descript records your screen, clones a voice for corrections (Overdub), writes with AI (Underlord), and lets a team comment on a project. That's a production suite. We do none of it — one file in, one cut out.

Multitrack and remote recording

Descript records and edits multiple synced tracks and remote guests. We take one mixed-down file. A two-mic podcast where you need per-mic control is a Descript job, not ours.

When Make Autocut is enough — and costs a fraction

The flip side. These are the jobs where paying a monthly subscription to a desktop suite is paying for ninety features to use one.

You just want the dead air cut, once

A one-take podcast or a talking-head YouTube video is mostly “remove the pauses and tighten it up.” On our public bench (2026-07-16) a long-form talk gave back 2.9% of its length in 10 razor cuts, a recorded lecture 1.4% in 7. That's the job — and it doesn't need a monthly subscription to a suite you open twice a month.

You can't (or won't) install software

Descript is a desktop app. On a managed work laptop, a Chromebook, or a borrowed machine you often can't install anything. This runs in a browser tab. Nothing registers, nothing needs admin rights, nothing to uninstall afterwards.

Your language isn't English

Descript transcribes, so its quality follows the language its speech model knows best. We never read the words — sound above the noise floor is kept, silence below it is cut — so French, Japanese, Arabic, a whistle or a dog bark are all treated the same. Nothing here is limited to the languages a model was trained on.

You care where the file goes

Descript keeps your project in its cloud until you delete it. We take the upload, cut it, and delete it automatically about an hour later — no account, no library, nothing kept for training. If the recording is sensitive, that difference matters more than any feature.

Every percentage above is from our public benchmark, not a marketing figure — the auto cut audio page has 155 threshold measurements on five downloadable recordings, including one where the tool honestly cuts nothing. A prepared voice-over, for reference, only gives back 3.3%.

Migrating from Descript, in four steps

There’s nothing to export from Descript in a special format and nothing to install here. You bring the same source file you brought to Descript, and leave with a cut file or an editable timeline.

  1. 1. Get the original file out

    You don't need to export anything special from Descript. Use the source you dropped into it — the raw MP4, MOV, WAV or MP3 you recorded. If you only have the Descript project, export a full-quality video or audio from it first (File → Export), then use that.

  2. 2. Drop it in — nothing to install

    Upload it here. Free tier takes up to 5 minutes and 200 MB per file; Pro goes to 60 minutes and 2 GB. Accepted in: MP4, MOV, MKV, WEBM, AVI, MP3 and more. It uploads straight to storage — nobody watches it and it's deleted about an hour later.

  3. 3. See the cut before it happens

    This is the part a black box hides. We read your file's own noise floor, set the threshold to mean − 5 dB, and show it on the waveform. Drag it, raise the minimum silence to keep thinking pauses, adjust the padding so joins don't clip — then cut. You approve the decision instead of trusting it.

  4. 4. Take the file, or take the timeline

    Download the cut MP4/MP3 if you just want the clean result. Or export an editable timeline — Premiere XML, FCPXML, EDL or OpenTimelineIO — and finish the edit in your NLE with every cut still reversible. No render, no re-encode; your media is never re-opened.

Editing in Premiere, Final Cut or Resolve after the silence pass? The Premiere, Final Cut and Resolve guides walk the timeline round trip click by click. Weighing up the cloud tools too? The Cleanvoice comparison puts ten silence tools in one table.

Cut a file and see the difference

First one is free, no account, no install. You see the waveform and the threshold it picked before anything is cut — then take the clean file, or take the timeline for your editor.

Drop your audio or video file
or browse
MP3, WAV, M4A, MP4, MOV, MKV, WEBM — up to 200 MB

For creators who record in one take: podcasters and YouTubers have their own walkthroughs.

Questions people actually ask

Is Make Autocut free, or is it a trial that expires?

The free tier is genuinely free, not a countdown. One file a day, up to 5 minutes and 200 MB, no account, no watermark, every export format unlocked. The honest limit is that 5 minutes won't hold a full podcast episode — if this becomes part of how you edit you'll move to Pro (60 min, 2 GB, no daily cap). But unlike Descript there's no monthly subscription to a whole suite just to cut silences.

Does it work offline like the Descript desktop app?

No — it's a web tool, so it needs a connection to upload and process. That's the trade for having nothing to install: Descript runs offline once installed but costs you a download, an install, updates and disk space; we cost you an upload. If you specifically need offline editing on a plane, the desktop app wins that one and we won't pretend otherwise.

Is the audio quality preserved when I cut silences?

If you take the timeline export, yes — nothing is re-encoded. The timeline is just a cut list your editor applies to your original file, which plays off your own disk untouched. The MP4/MP3 export does re-encode (H.264 CRF 20, or LAME -q:a 2, transparent for speech), so if you're protecting a master, take the timeline and skip the render button entirely.

Does it remove filler words like Descript does?

No, and this is the honest dividing line. Descript removes “um”, “uh” and repeated words because it edits a transcript. We detect silence by loudness and have no transcript, so verbal tics — which are sound, not silence — survive. If filler removal is your actual need, use Descript. If it's the pauses and dead air, that's us, and you don't need the rest of the suite to do it.

What happens to my file? Is it GDPR-friendly?

Your upload is deleted automatically about an hour after processing, it's never used to train anything, and the free tier needs no account, so there's no profile attached to it. Descript keeps your project in its cloud library until you delete it — fine for a collaborative workflow, more than you want for a one-off sensitive recording. For a single interview you'd rather not leave lying in someone's cloud, upload-cut-delete is the cleaner path.

Can I still finish the edit in a real editor afterwards?

Yes — that's what the timeline export is for. Cut the silences here, download a Premiere XML, FCPXML, EDL or OpenTimelineIO file, and every removed pause lands as a straight cut you can slide, ripple or undo in Premiere, Final Cut, Resolve or anything that reads OTIO. You get the fast silence pass without giving up hand control of the final edit.

Try it on a real file. No account, no install, no card.

5 minutes, free, no watermark — every export format, timeline included, unlocked.

Cut my file free