How to Use
ZoopZip is designed to be obvious: drop a file, pick a model, and get your transcript. Here's the full walkthrough of FREE offline speech to text, plus what every option does.
Open Configuration → Manage Models and download one. Start with small — it's a great balance of speed and accuracy. This is the only step that needs the internet.
Drag files onto the window, or click Choose Files. ZoopZip accepts MP3, M4A, WAV, MP4, MOV, MKV and more — and everything stays on your computer. You can also paste a YouTube link instead of a file.
Choose the model, the spoken language (or leave it on auto-detect), and which formats to save. See options explained below.
ZoopZip works entirely offline and writes the transcript next to your file — or into a folder you choose. That's FREE offline speech to text, start to finish.
Sets which Whisper model does the transcribing. Bigger models are more accurate but slower and use more memory. A ⬇ mark means a model still needs downloading; unmarked models are ready to use. Start with small — but for non-English audio, consider large-v3 for much better accuracy.
The language spoken in your file. Leave it on Detect language unless the guess comes out wrong — naming the language explicitly can improve accuracy for tricky recordings.
Click the YouTube toggle at the top of the window to paste a video link instead of dragging a file. ZoopZip downloads the video (up to 1080p, one at a time) and adds it to the queue automatically. Full details on the YouTube transcription page.
Tidies the transcript before it's saved: filler words (“um”, “uh”, “you know”) removed, immediately repeated words collapsed, and spacing, punctuation and capitalisation smoothed out. Timings are untouched, so subtitles stay in sync. Runs on your computer — nothing is sent anywhere.
Tags each segment with Speaker 1, Speaker 2 and so on, so interviews and conversations read as a dialogue. Labels appear in every format you save. ZoopZip works this out from the pauses between turns rather than from the sound of each voice, so treat it as a readability aid: speakers who talk over each other, or who don't pause between turns, may be labelled wrongly.
Under Configuration → Live Transcribe…. Records your microphone and transcribes it in a rolling few-second window, so text appears while you're still talking. You can edit the transcript as it's being written, and save it anywhere with Save As…. Choose the language before you start — over a few seconds of audio, naming the language is noticeably more reliable than letting it detect.
Plain text, just the words with no timings. Best for notes, documents and copy-paste.
SubRip subtitles: the words plus when each line is spoken. Use these for video subtitles on YouTube, VLC, Premiere and CapCut.
WebVTT subtitles — the same idea as SRT, in the format web video players expect. Choose this for a website.
Every segment with its exact start and end time, for feeding into other software. For developers.
By default transcripts are written beside the original file. Choose Configuration → Save Location → Choose Folder to send them all to one folder instead — and create a new folder right there if you need one.
The language of the ZoopZip interface itself (separate from the audio language). Found under Configuration → App Language, with 12 languages available.
Not sure which to pick? Use this quick guide — and see the full system requirements for memory and speed per model.
| Feature | Free | Pro (one-time) |
|---|---|---|
| Offline transcription & privacy | ✓ | ✓ |
| Files queued | One | Unlimited, run one after another |
| Models | up to medium | + large-v3 |
| Export formats | TXT, SRT | + VTT, JSON |
Yes. ZoopZip is FREE offline speech to text — the core app costs nothing and needs no account. An optional one-time Pro upgrade unlocks batch transcription, the large-v3 model and extra export formats.
No. Everything happens on your computer. Your audio is never uploaded, and once a model is downloaded ZoopZip works with no internet connection at all.
Only once, to download a speech model. After that ZoopZip transcribes completely offline.
Audio and video: MP3, M4A, WAV, FLAC, OGG, MP4, MOV, MKV, AVI, WEBM and more. ZoopZip extracts the speech automatically.
Plain text (TXT) and SubRip subtitles (SRT) are free. WebVTT (VTT) and JSON are available in ZoopZip Pro.
ZoopZip transcribes many languages including English, Thai, Chinese, Japanese, Korean, Spanish, French, German and more, and can auto-detect the spoken language. For non-English audio, the large-v3 model gives noticeably better accuracy than the smaller models.
Yes — it's a lifetime upgrade. Pay once, and every future ZoopZip Pro update is included at no extra cost. Yours forever, on your computers.
ZoopZip