Lightning Notes vs Whisper文字起こし

Record. Transcribe. Summarize. All on your device. Transcription and speaker separation run entirely on your iPhone — no audio ever leaves your device. Stop the recording and on-device AI works out who said what, producing a speaker-labeled transcript. Then choose local AI or Cloud AI to write the meeting notes. Supported by ads, with all core features available without a subscription. ------------------------ Key Features ------------------------ ■ Accurate Transcription — On-Device Apple's speech recognition engine processes your audio entirely on-device. Meetings, lectures, interviews — transcribed without sending audio anywhere, and reliable even for long recordings. ■ Automatic Speaker Separation Speaker separation runs automatically on your iPhone once recording stops, identifying who said what across multi-participant conversations. Voice fingerprints are learned over time, so names get assigned automatically in later meetings. Manual editing is always available. ■ AI Meeting Notes — Your Choice of AI · On-device AI: No internet required; nothing leaves your iPhone. Needs 6 GB RAM or more (iPhone 13 Pro and later). · Cloud AI: DeepSeek V4-Flash via OpenRouter, streamed in real time. API data is not used for model training, and only Zero Data Retention (ZDR) endpoints are used, so the provider does not store your data to disk. ■ Per-Speaker Summary (Cloud AI) A breakdown for each participant: role, key points raised, and behavior during the meeting. See who drove decisions and who needs follow-up. ■ Person Profiles (Cloud AI) Cross-reference up to 30 past meetings featuring a person and generate an AI profile of their role, communication style, and project involvement. ■ Dictionary Correction (Custom Terms) Register company names, product names, and jargon, and the app fixes where speech recognition misheard them — "Tower Electric" becomes "Towa Electric". Rather than swapping on sound alone, an on-device language model checks the replacement reads naturally in context, so unrelated passages are left alone. English and Japanese supported. Corrections are listed with before/after text, can be reverted one at a time, and apply retroactively to past meetings. ■ Audio & Video File Import Import external audio or video recordings and run them through the same transcription and speaker separation pipeline. ■ Share & Export Export as PDF, choosing which sections to include (summary, speaker summaries, transcript). Speaker colors and tables are preserved. Text and audio sharing are also supported. ■ Privacy-First Architecture Transcription, speaker separation, and local summarization all run on-device. Cloud AI is opt-in — leave it off and no recording content ever leaves your iPhone. ■ Folder Organization Organize meetings into folders by project, client, or topic. ■ Full-Text Search Search titles, summaries, and transcripts. Find any meeting by keyword in seconds. ■ Edit Transcripts Fix recognition errors directly in the app and save an accurate record. ■ Control From the Lock Screen Elapsed time appears on the Lock Screen and Dynamic Island while recording, with pause, resume, and stop right there. Add the Control Center button to start recording without launching the app. ■ Backup & Restore Export meetings, transcripts, and speaker data to a file and restore it on another device or after reinstalling. You keep the file; nothing goes to an external server. ■ Meeting Types · Meeting: Multi-participant discussions · Lecture / Seminar: One-to-many presentations or classes · Interview: One-on-one or small group conversations ------------------------ Requirements ------------------------ · Transcription, speaker separation, file import: iPhone 12 or later (A14 chip), or iPad with M1 or later · Local AI summarization: 6 GB RAM or more (iPhone 13 Pro, iPhone 14 series, iPhone 15 and later) · Audio/video formats: WAV, M4A, MP3, MP4, and other common formats --- Take back your meetings. No subscription required.

Whisper文字起こしは、OpenAI Whisper（whisper.cpp）を iPhone 上で動かす音声文字起こしアプリです。 【主な機能】 ・マイク録音の文字起こし ・音声・動画ファイルのインポート（m4a / wav / mp3 / mp4 / mov など） ・多言語（日本語、英語、中国語ほか設定画面の言語一覧） ・履歴の保存・検索・お気に入り ・txt / json / csv / srt へのエクスポート、共有 【オフラインについて】 初回起動時に Whisper モデルを Hugging Face からダウンロードします（Base 既定で約142MB）。ダウンロード完了後、文字起こし処理は端末内のみで完結し、音声・文字起こし結果を開発者サーバーへ送信しません。 【必要環境】 ・iOS 17.0 以降の実機（Simulator 非対応） ・マイク利用時は許可が必要です 広告・トラッキング・アプリ内課金はありません。 【免責】文字起こし結果は AI による参考情報であり、正確性は保証されません。重要な用途では必ず内容をご確認ください。 https://porarrirr.github.io/offline-whisper-transcriber-ios/disclaimer.html