Cut your podcast by editing the transcript
Tighten an episode the way you'd edit a doc: it transcribes your recording on-device, and you cut the audio by deleting words. Drop a tangent, strip every "um," and export a clean WAV. It's a text-based cutter — not a full DAW — free, in your browser, nothing uploaded.
Replaces tools like
Works well with
Eziclip is a free, 100% in-browser way to CUT a podcast by editing its transcript: drop your episode — MP3, M4A, WAV, or a video (it uses the audio) — an on-device AI transcribes it, and you trim the audio by deleting words: drop a flub, a tangent or a whole answer and that exact slice is gone, or strip every filler ("um," "uh") and silence in one tap. It's a text-based cutter, not a full podcast editor — there's no multitrack, music or mixing — so you take the clean, lossless WAV (channels intact) on to your final mix. Every cut is non-destructive and reversible. Nothing is uploaded, with no account, no watermark and no length cap.
Edit your podcast the way you'd edit a doc
Editing a podcast on a waveform is slow and blind — you're hunting for the right squiggle, scrubbing back and forth to find where someone restarted a sentence, nudging selection handles a few milliseconds at a time. It turns that on its head. The moment your episode loads, an on-device AI has transcribed it, and you're reading your whole conversation as text. You edit the audio by editing the transcript: it's the difference between hunting for a sound and finding a word.
This is built for the work podcasters actually do — tightening an episode after the recording is done. You read the conversation top to bottom, see the flubs, the false starts and the tangents in plain text, and cut them where you spot them. No timeline scrubbing, no learning a DAW. If you can edit a paragraph, you can edit your episode.
Delete a word, lose that exact slice of audio
Highlight a stumbled word, a repeated phrase, a flubbed line, or a whole three-minute tangent that didn't earn its place — and that exact slice of audio is cut out. The transcript and the sound stay locked together, so deleting text is deleting audio. Trim a guest's rambling answer down to the sharp version, drop the part where someone's phone rang, lift out the bit you promised to keep off the record — all by selecting and deleting, the same gesture you'd use in any text editor.
Nothing here is one-way. Every edit is non-destructive: the audio you cut isn't gone, it's set aside, and you can restore any word, line or section you change your mind about. The cut itself is reversible too — step it back like an undo — so you can edit aggressively, tightening hard, knowing the original take is always one move away. That safety net is what lets you cut a podcast down to its best self instead of timidly nibbling at the edges.
Strip every "um," "uh" and dead-air gap in one tap
Two things bloat almost every raw episode: filler words and silence. It handles both in a single tap. One button strips every filler — the "um," the "uh," the reflexive throat-clearing that pads a conversation — across the entire episode at once, instead of you hunting each one down. Another clears the silences and dead-air gaps: the long pauses, the leading and trailing dead air, the breath-holding moments between thoughts that make an episode drag.
This is the pass that turns a loose, conversational recording into something that moves. A tighter episode respects your listener's time, and removing the filler and the dead air is most of what "good editing" sounds like. Because it's all non-destructive, you can run the one-tap clean-up, read the result, and restore any pause that was actually doing dramatic work — the deliberate beat before a punchline, the silence that lets a heavy moment land.
Your recording never leaves your device
The whole thing runs on your own machine. Your episode is decoded and transcribed in your browser, and every edit and the final export happen there too — there's no upload, no server processing your audio, and nothing for anyone to keep. That's not a privacy promise written by a lawyer; it's how the tool is built. An unreleased episode, a sensitive interview, a guest conversation you haven't cleared yet — none of it touches a cloud, because there's nowhere for it to go.
For podcasters that matters more than for most. You're often sitting on an embargoed interview, a sponsor read that isn't public, or a candid conversation a guest trusted you with. Here, it stays between you and your laptop. The AI transcription model downloads to your browser once, then everything runs locally — and keeps working offline once the page has loaded.
Free, no catch — clean WAV, channels intact
This is free for every podcaster, and free is a choice about how the tool should work, not a stripped-down trial of something better. There's no account to create, no watermark stamped on your audio, and no length paywall — a three-hour interview edits the same as a ten-minute solo episode, with no cap waiting to upsell you. You get the full tool every time.
When the episode reads the way you want, you export a lossless WAV — no re-compression eating your audio quality, with the channels preserved so a stereo or two-mic recording exports the way it came in, ready for your final mix or straight to publish. The download is yours, full quality, no sign-up. If the tool saved you an hour of waveform editing and you want to chip in, there's a button — but the clean WAV is yours either way.
How to edit a podcast by text in three steps
- 1
Drop your episode in
Add an MP3, M4A or WAV — or a video file, and it uses the audio. An on-device AI transcribes it right in your browser; nothing is uploaded to a server. In a moment you're looking at your whole episode as an editable transcript.
- 2
Cut by text, or clear filler and silence in one tap
Delete a word, a flubbed line or a whole tangent in the transcript and that exact slice of audio is cut out. Or hit one button to strip every filler word — "um," "uh" — and every silence and dead-air gap across the episode. Every cut is non-destructive and reversible, so restore anything you change your mind about.
- 3
Export a clean WAV
When the episode reads tight, export a lossless WAV with the channels preserved — ready to drop into your publishing flow or a final mix. Free, no watermark, no sign-up, no length limit.
Questions
The transcript and the audio are locked together. When you delete a word, a line or a whole section of text, the exact slice of audio under it is cut out — so editing the episode is just editing the words. It works the way a podcast text editor should: you read the conversation, highlight what doesn't belong, and delete it. Every cut is non-destructive and reversible, so you can restore anything you change your mind about.
Yes — that's a one-tap pass. One button strips every filler word, the "um" and "uh" across the whole episode at once; another clears the silences and dead-air gaps, including the leading and trailing dead air. It's non-destructive, so after the clean-up you can read the result and restore any pause that was doing real work, like a deliberate beat before a punchline.
Drop in an MP3, M4A or WAV — or a video file, and it uses the audio. You get back a clean, lossless WAV with the channels preserved, so a stereo or two-mic recording exports the way it came in, with no re-compression. It's ready to drop into your final mix or publish as-is.
No. Everything — the AI transcription, every edit, and the final export — runs on your own device in the browser. There's no upload and no server touching your audio, so an unreleased episode, an embargoed interview or a private guest conversation never leaves your machine. The transcription model downloads to your browser once, then it works offline.
No. It's free with no account, no watermark, and no length paywall — a three-hour interview edits the same as a short solo episode. Free is a deliberate choice here, not a limited trial: you get the full editor and a full-quality WAV export every time.