The app is now Verbatim, with clearer private voice-to-text positioning throughout the experience. This update also refreshes Voice Keyboard guidance, clarifies on-device processing and English translation, and keeps the same recordings, settings, trial state, and Verbatim Pro purchase. Limited time offer to go pro for $1.99. No monthly fees.

Verbatim-AI 1.4.0 introduces the Verbatim voice keyboard, so you can dictate into supported apps with private, on-device transcription. Also new: • Unlimited access to every feature during a 3-day free trial • Faster preparation for dictation and transcription • Improved keyboard setup and first-use guidance • More reliable background dictation and text insertion • Privacy and stability improvements

Version 1.3.0 brings a clearer free-to-Pro experience, improved App Store screenshots, and a more polished on-device AI workflow. New in this update: • Refined AI Search experience • Improved AI chat/search positioning • Better action feedback across key buttons • Polished live captions and translation workflows • General UI and usability improvements

What’s New in Version 1.2.1 Now free to try. Record and transcribe instantly with 3 recordings and 1 AI summary included. Upgrade to Pro to unlock unlimited recordings, smarter summaries, and AI search across your transcripts. You can now import audio and video files directly into Verbatim-AI for transcription, making it easier to work with meetings, interviews, lectures, voice notes, podcasts, and video recordings you already have saved. Supported import types include audio and video files recognized by iPhone Files and Photos, including common formats such as MP3, M4A, WAV, AAC, MP4, and MOV. New in this release: - Import audio files for transcription - Import video files for transcription - Import videos from Camera Roll - Translate supported non-English recordings into English - Detect and display the spoken language in English - Preserve speaker labels when translating multi-speaker recordings (Beta) - Improved transcription, translation, and summary actions with a cleaner and more intuitive design Verbatim-AI supports transcription in 99 languages: Afrikaans, Albanian, Amharic, Arabic, Armenian, Assamese, Azerbaijani, Bashkir, Basque, Belarusian, Bengali, Bosnian, Breton, Bulgarian, Burmese, Catalan, Chinese, Croatian, Czech, Danish, Dutch, English, Estonian, Faroese, Finnish, French, Galician, Georgian, German, Greek, Gujarati, Haitian Creole, Hausa, Hawaiian, Hebrew, Hindi, Hungarian, Icelandic, Indonesian, Italian, Japanese, Javanese, Kannada, Kazakh, Khmer, Korean, Lao, Latin, Latvian, Lingala, Lithuanian, Luxembourgish, Macedonian, Malagasy, Malay, Malayalam, Maltese, Maori, Marathi, Mongolian, Nepali, Norwegian, Nynorsk, Occitan, Persian, Polish, Portuguese, Punjabi, Pashto, Romanian, Russian, Sanskrit, Serbian, Shona, Sindhi, Sinhala, Slovak, Slovenian, Somali, Spanish, Sundanese, Swahili, Swedish, Tagalog, Tajik, Tamil, Tatar, Telugu, Thai, Tibetan, Turkish, Turkmen, Ukrainian, Urdu, Uzbek, Vietnamese, Welsh, Yiddish, Yoruba.

What’s New in Version 1.2.0 You can now import audio and video files directly into Verbatim-AI for transcription, making it easier to work with meetings, interviews, lectures, voice notes, podcasts, and video recordings you already have saved. Supported import types include audio and video files recognized by iPhone Files and Photos, including common formats such as MP3, M4A, WAV, AAC, MP4, and MOV. New in this release: - Import audio files for transcription - Import video files for transcription - Import videos from Camera Roll - Translate supported non-English recordings into English - Detect and display the spoken language in English - Preserve speaker labels when translating multi-speaker recordings (Beta) - Improved transcription, translation, and summary actions with a cleaner and more intuitive design Verbatim-AI supports transcription in 99 languages: Afrikaans, Albanian, Amharic, Arabic, Armenian, Assamese, Azerbaijani, Bashkir, Basque, Belarusian, Bengali, Bosnian, Breton, Bulgarian, Burmese, Catalan, Chinese, Croatian, Czech, Danish, Dutch, English, Estonian, Faroese, Finnish, French, Galician, Georgian, German, Greek, Gujarati, Haitian Creole, Hausa, Hawaiian, Hebrew, Hindi, Hungarian, Icelandic, Indonesian, Italian, Japanese, Javanese, Kannada, Kazakh, Khmer, Korean, Lao, Latin, Latvian, Lingala, Lithuanian, Luxembourgish, Macedonian, Malagasy, Malay, Malayalam, Maltese, Maori, Marathi, Mongolian, Nepali, Norwegian, Nynorsk, Occitan, Persian, Polish, Portuguese, Punjabi, Pashto, Romanian, Russian, Sanskrit, Serbian, Shona, Sindhi, Sinhala, Slovak, Slovenian, Somali, Spanish, Sundanese, Swahili, Swedish, Tagalog, Tajik, Tamil, Tatar, Telugu, Thai, Tibetan, Turkish, Turkmen, Ukrainian, Urdu, Uzbek, Vietnamese, Welsh, Yiddish, Yoruba.

What’s New: **Transcription** - Replaced the old transcription path with on-device WhisperKit transcription. - Switched from the English-only Whisper model to the multilingual `openai_whisper-small` model. - Bundled speech models locally so transcription runs fully offline. - Improved transcript cleanup so Whisper control tokens and blank-audio markers are removed from saved transcripts. **Speaker Diarization** - Added on-device speaker diarization using SpeakerKit. - Speaker-labeled transcripts now display as `Speaker A`, `Speaker B`, and so on. - Added a setting to make speaker diarization optional. - Added speaker count controls with options like `Auto`, `2 speakers`, `3 speakers`, and `4 speakers`. - Moved transcription/speaker settings into a cleaner gear-based settings flow on the recording screen. **Summary Generation** - Reworked the summary prompt to keep summaries grounded in the actual transcript and reduce invented details. - Improved handling for devotional, lyrical, poetic, narrative, news, and legal-style content. - Reduced false action items by only allowing them when the transcript actually supports them. - Improved long-summary reliability with: - chunk checkpointing - partial summary persistence - automatic resume after reopening the app - fallback handling when a chunk returns malformed structured output - Increased summary context and reduced over-chunking for better quality on long recordings. - Updated the summary progress screen with clearer messaging that longer summaries may take several minutes. **Offline / Reliability** - Bundled speech model assets for fully offline transcription and diarization. - Kept the screen awake while summary generation is actively running. - Added resume support so long summaries can continue from the last completed chunk instead of starting over. **App Quality** - Cleaned up Swift concurrency issues in the app layer. - Kept existing recording, transcript, and summary flows working with the new speech pipeline. - Preserved a production-friendly offline architecture with no cloud dependency for transcription or diarization.

What's New in Version 1.0.2 • Improved transcript summaries to be more accurate, grounded, and faithful to the actual recording. • Fixed an issue where long recordings could lose the last few seconds during transcription. • Upgraded AI Search so natural-language queries work better and matching results are easier to open and review. • Improved summary progress feedback with a clearer loading experience. • Refined the recording detail view and added export options for transcripts and summaries. • Added a new Legal Info screen and additional app polish throughout. • Minor bug fixes and performance improvements

What’s New in Version 1.0.1 • Improved AI Search functionality • Users can now search recordings more naturally (e.g., “Find recordings about Amazon”) • Minor bug fixes and performance improvements