Dictofon

Capture an idea before it disappears. Dictofon records your voice, turns speech into editable text, and keeps recordings, transcripts and translations together.

RECORD AND TRANSCRIBE • Start recording with one tap and keep recording while you use another app. • Transcribe saved recordings offline with the bundled NVIDIA Parakeet model. It supports 25 European languages, including English, Russian, German, French, Spanish, Italian, Portuguese, Polish and Ukrainian. • See a live transcript when Apple's on-device speech recognition supports your device and selected language. Apple speech models may need an initial download. If live transcription is unavailable, you can still record and transcribe afterward. • Capture recordings on Apple Watch and transfer them to your iPhone.

FIND AND REFINE • Organize notes with folders, colored tags and favorites. • Search titles and transcripts, filter recent notes, and recover notes from Trash. • Listen with playback speed controls and 15-second skips. • Jump from timed transcript words to the corresponding audio position. • Edit transcript text, add bookmarks, trim recordings, remove a selected section and transcribe edited audio again.

TRANSLATE AND SHARE • Translate transcripts with Apple's on-device Translation framework. Supported language pairs require language packs; download them once before using translation offline. • Clean up text, adjust its tone, create lists or summarize with Apple Intelligence on supported devices and languages, when enabled and available. Basic text cleanup is also available without Apple Intelligence. • Export text, PDF, SRT or VTT subtitles, or share your audio. • Start recording and quick dictation through Shortcuts, Siri or the Action Button. Quick dictation can copy the result to your clipboard.

YOUR NOTES, YOUR CONTROL No Dictofon account, advertising or analytics. Speech recognition runs on your device. Optional iCloud synchronization uses your personal iCloud account; notes and recordings can sync through Apple's services. Manage iCloud access in iOS Settings. Sharing or exporting sends content only where you choose.

Requires iOS 26 or later. The interface is available in English, Russian and Simplified Chinese. The bundled speech model supports 25 European languages; Chinese, Japanese and Korean speech recognition are not offered in this version. Apple speech, translation and Apple Intelligence availability varies by device, language and system settings.

Speech recognition powered by NVIDIA Parakeet TDT 0.6B v3 (CC-BY-4.0). Core ML conversion by FluidInference; on-device runtime FluidAudio (Apache 2.0).