WhisText

About

Just tap record for instant transcription. Summaries and meeting minutes can be generated with a single tap. It also supports importing audio files and OCR transcription from camera/photos. Generated content can be posted automatically to Slack. Webhook integration lets you send results to Discord, n8n, Make, Zapier, and virtually any other service (Header authentication supported). Transcription is performed on-device using Apple Speech, while AI processing such as summaries and minutes runs on the server. Main Features - Tap record for instant transcription - Speaker diarization (powered by NVIDIA Nemotron 3; automatically identifies multiple speakers and color-codes them; can be toggled in Settings) - Live Preview: see real-time transcription while recording - Automatic summary and meeting minutes generation via AI (custom prompts supported) - Webhook integration (send to Discord, n8n, Make, Zapier, and more — just set a URL; Header authentication with an API key is also supported) - OCR transcription from images (batch processing for multiple images) - Automatic posting to Slack - Insert timeline markers to flag key moments on the spot (adjustable in 5-second increments) - Transcript highlighting synced to playback position Fast, Multilingual Processing - Apple Speech provides fast, on-device transcription for supported languages. - Supported languages include Japanese, English, Chinese, Korean, French, German, Italian, Spanish, Portuguese, and Cantonese. - Server-side processing is used for AI tasks such as summaries, minutes, and titles. Recording & Import - Tap to record → instant transcription - Import audio files (Share Sheet supported) - Automatically extract and compress audio from video files - Import images for OCR (processed on-device) Export & Integration - Export as audio, text, summary, minutes, or subtitles (VTT, SRT) - Automatic posting to Slack - Webhook integration (Discord, Slack, n8n, Make, Zapier, and more) Privacy & Security - Audio files are not sent to the server for transcription. - Transcripts are sent to the server for AI processing such as summaries and minutes. - Server-side processing data is promptly deleted once the AI task completes. Supported Formats - Audio: mp3, m4a, aac, wav, flac, ogg, opus, wma, amr, mpga, webm, aiff, caf - Video: mp4, mov, m4v, webm, mkv, avi, mpeg, mpg *Recognition accuracy may vary depending on audio quality and content.