VoiceKit: Filler Cut & Subs

VoiceKit removes filler words like “um” and “uh” from your video and audio, generates subtitles, and summarizes and translates what was said. Transcription, filler detection, subtitles, summaries, and translation all run on your iPhone. Your voice never goes to the cloud. ■ Three tools, one app • Filler removal — hesitations are detected for you, and you decide what gets cut • Subtitles — export SRT from the transcript, or burn subtitles into the video (Pro) • Summaries — turn a meeting or lecture recording into an overview and key points (Pro) On top of that, translate transcripts and summaries on device with Apple’s Translation, and export translated SRT files that keep the original timestamps. ■ Record right here — even with the screen locked Record a meeting, a lecture, or a piece to camera inside the app and go straight to processing — no file shuffling. Audio recording keeps going when you lock the screen or switch to another app. You can also import from your photo library or the Files app. ■ Results saved to History automatically Transcripts, summaries, translations, and audio recorded in the app are saved without pressing a save button. Reopen them any time; summaries and translations you already made appear instantly. History stays on your device only, and you can swipe to delete any item. ■ Detected automatically, finished by you Detection is automatic, but nothing is deleted on your behalf. Play each candidate and keep it or reject it. For the ones automation cannot get — a quiet “um” that runs into the word before it — mark IN and OUT on a large scrubber and cut it yourself. Automation does the prep; your ears do the finish. ■ Not one byte goes to the cloud Speech recognition, filler detection, subtitles, summaries, and translation are all processed on your device. Summaries use Apple’s on-device foundation model — no cloud, not even Private Cloud Compute. No account, no tracking, no ads. Note: the network is used only to download the speech recognition model on first launch (about 626 MB) and, when you translate into a language for the first time, to let iOS download that language. Only the download request leaves your device — never your audio or text. ■ One-time purchase VoiceKit Pro is a one-time purchase. It is not a subscription. WHAT'S FREE • Transcription (Japanese and English) • Filler detection and cutting • SRT subtitle export • Transcript translation and translated SRT export • History (automatic saving of results and recordings) • Manual editing (up to 10 edits) Free processing is limited to 5 minutes per file, and exported video carries a watermark. VOICEKIT PRO (ONE-TIME PURCHASE) • No length limit • Unlimited manual edits • No watermark • Burn subtitles into video • Summaries (and summary translation) • Filler detection sensitivity control • Custom dictionary for your own filler words ■ Made for • YouTubers and short-form creators tightening up their delivery • Online instructors adding subtitles to lessons and training videos • Professionals turning meeting recordings into notes • Students recording lectures to review later • Podcasters cleaning up a take quickly ■ Requirements • iPhone (iPad is not supported), iOS 17.0 or later • Summaries require iOS 26 or later on an Apple Intelligence–capable device (iPhone 15 Pro or newer) • Translation requires iOS 18 or later (a language used for the first time is downloaded by iOS) • An internet connection and about 1 GB of free space for the one-time model download Support: https://hibinokou-spec.github.io/voicekit-support/ Privacy Policy: https://hibinokou-spec.github.io/voicekit-support/privacy.html