Guide
Transcribe a video on iPhone, privately
Updated
CaptionAddr transcribes the English speech in a video on your iPhone or iPad itself, using a speech model built into the app, so your video, audio and transcript aren’t sent to CaptionAddr’s servers for transcription. The transcript becomes editable, word-by-word captions that you can burn into a new copy of the video.
How on-device transcription works
CaptionAddr includes an English speech-recognition model, NVIDIA’s Parakeet TDT 0.6B v2, in the app download. When you generate captions, your iPhone runs that model on the audio of your local copy of the video. Nothing needs to be uploaded for the words to appear, and there’s no account to sign in to.
That makes CaptionAddr a good fit for videos you’d rather keep on your device: family clips, client work, unreleased content or anything personal.
What works offline
- Works without internet
Transcribing a video that is already on your device, editing captions, styling and exporting a captioned movie.
- May need internet
Downloading an original from iCloud Photos, buying or restoring CaptionAddr Pro, verifying a subscription, and sharing to online services.
- Other online services
Separately from transcription, the app uses online services for purchases, diagnostics, security and aggregate workflow counts. What each one collects is listed in the privacy policy.
Transcribe a video privately
- Make sure the video is on your device. If it’s in iCloud Photos, open it in Photos once so the original downloads.
- Open it in CaptionAddr. Tap Select Video and choose the clip. CaptionAddr works with a local copy and leaves the original alone.
- Let the captions generate. Captions start as soon as the video opens; keep the app open while the on-device model transcribes.
- Review the transcript. Read through the words in Edit Captions and correct anything misheard.
- Export if you want a captioned video. Save a new movie with the words burned in to Photos.
Is CaptionAddr right for your transcription?
CaptionAddr turns speech into on-screen captions. It doesn’t export the transcript as a text document, SRT or VTT file, so if you need a written transcript to paste into a document, a dedicated transcription tool will suit you better. It supports English speech only, with no translation. Free covers videos up to 2 minutes and CaptionAddr Pro covers up to 30 minutes. Like any speech model, it can make mistakes with noise, music, overlapping voices or strong accents.
Frequently asked questions
Can I transcribe a video on iPhone without uploading it?
Yes. CaptionAddr transcribes English speech on the device with a bundled model, and doesn’t send your video, audio or captions to its servers for transcription.
Does CaptionAddr work in airplane mode?
Transcription, editing and export work without a connection once the video is on your device. Buying or restoring Pro, Pro verification and iCloud downloads need internet.
Can I export the transcript as text?
No. CaptionAddr exports a video with captions burned in. It doesn’t export text, SRT or VTT files.
Which languages can it transcribe?
English only. The bundled model is designed for English speech.
How accurate is it?
Accuracy is generally good for clear English speech and drops with background noise, music or several people talking at once. Always review the words before sharing.