Join the waitlist
Early access · English version in development

Turn full podcast episodes into clips without uploading them

OxClipper transcribes the episode on your own computer, has AI propose several clip versions from the transcript, and lets you check every segment before you batch-export. The episode file stays on your machine, and local work isn't metered by the minute.

It's in use in Chinese today. The English version isn't out yet, so this page is a waitlist, not a download.

We'll email you when the English version opens. No spam, no sharing. Privacy

Why podcasters look for another way

Full episodes are long. Upload-and-meter tools feel it most.

Uploading takes a while

A 90-minute video episode is a big file. Here there's nothing to upload: you point the app at the file on your disk.

Minute quotas run out

A weekly 90-minute show is about 390 minutes a month. Local transcription and rendering here aren't counted by the minute; AI clip planning is billed by transcript length.

Unreleased episodes stay put

With the local engines, audio and video never leave your computer. Clip planning sends only the transcript text to an AI service.

One episode, start to finish

What a 90-minute episode looks like

This is how the current (Chinese) version works. The English version is planned around the same workflow.

  1. Drop it in

    Add the episode file, plus its SRT if you already have one. Several episodes can go in one project (up to 100 files).

  2. Transcribe locally

    Whisper is the engine to pick for English speech. The model downloads once; after that it runs on your computer.

  3. Get versions

    AI reads the transcript and proposes up to 10 clip versions per run. Run it again to add more; earlier ones stay.

  4. Check each segment

    Play every segment, nudge in/out points by 0.1 s or type exact times, remove or reorder. Changes save automatically.

  5. Batch export

    MP4s with captions and a title, as 9:16 vertical, landscape, or a custom canvas size. Up to 5 at a time.

Ways to cut an episode

Pick the kind of clip you want

Highlights

One idea, story or strong line per clip. Tell it what you're after in plain words, like "keep the advice and the reason behind it".

Topic compilation

Name a topic and it gathers the moments about it. Put several episodes in one project to pull a topic from across them.

Condensed

Drops repeated and rambling sentences and keeps the core, for a tighter cut of a segment or a whole conversation.

Straight talk

What it doesn't do

  • It doesn't tell the host and guests apart, so you can't ask for "only what the guest said".
  • On multi-camera shows, it won't switch to whoever is speaking, and there's no face-tracking reframe.
  • Captions are sentence-level. It won't remove "um" and "uh" inside a sentence.
  • AI picks clips from the transcript, not from the picture.
  • Transcription can mishear names and numbers, so check them before you publish.
  • The English interface isn't released yet. That's what the waitlist is for.

Comparing tools? See OxClipper as an Opus Clip alternative.

FAQ

Questions

Does it work with English-language podcasts?
The local Whisper engine transcribes English speech, and the transcription language can be set to English, Chinese, Japanese or Korean. The interface, presets and help are in Chinese today; bringing those to English is what the English version is for.
How long can an episode be?
There's no per-minute quota on local transcription and rendering. Each AI analysis run has a transcript length cap, roughly four to six hours of continuous speech, so a normal episode fits easily.
Do I need a GPU?
No. Export uses hardware encoding when your computer has it and falls back to the CPU otherwise. Speed depends on your hardware.
How much will it cost?
English pricing hasn't been set. Waitlist members hear about pricing, and any early-access offer, first.

Be first to try the English version

One email when it opens. That's it.

No spam, no sharing. Privacy