What's new in 2.0:

• Start a recording without opening the app. Ask Siri, press the Action button, use a Control Center control or a Home Screen widget — and while it runs, the elapsed time sits on the Lock Screen and in the Dynamic Island.
• Summaries can arrive on their own. Turn them on, and a recording that finishes analyzing comes with a recap already written.
• The app now tells you when a transcript is ready, when a recording lands from your Apple Watch, and when analysis fails. It had never posted a notification before.
• Calendar awareness. Let Synopsule read your calendar and it offers to record a meeting as it begins, names the recording after the event, and seeds the speaker count from the invitees. Read-only, on your iPhone, never written back.
• Import an audio file — a voice memo, a downloaded call, a colleague's recording — and transcribe it like anything you recorded yourself.
• Reviewing an hour is finally practical: 1.25x, 1.5x and 2x playback, 15-second skips, a transcript that follows the playhead, and a waveform drawn from the real audio instead of decorative bars.
• Correct a transcript afterwards. Notes and flags can be added, edited and deleted at any point — a mis-tapped flag used to be permanent — find and replace works across a transcript, and selection spans lines.
• Search reads transcript bodies now, so a recording you never summarized is still findable and an answer quotes what was actually said.
• Action items from every recording collect in one place, grouped by owner or due date, recurring meetings are recognized as a series, and the library can be sorted, filtered, tagged, and worked on many recordings at a time.
• Back up the whole library to a single file and restore it by merging, so nothing already here is overwritten. It is also how you move a library between an iPhone and a Mac.
• PDF export, audio export, renaming from the reader, re-running analysis on a finished recording, real storage sizes, and a delete you can undo.
• Setting up a speech model finally reports itself honestly: real progress with the stage it is on, a Cancel that keeps what has already downloaded so starting again resumes, and no silent restart of a download you stopped.
• The whole app follows your text size, touch targets meet 44pt, Reduce Motion is honored, and the speaker palette was rebuilt so speakers stay distinguishable for color-blind readers.

On Apple Watch:

• The status on your wrist is now the truth. Delivered, importing and imported are distinct states, and each is claimed only once it has actually happened.
• A transfers list shows every recording waiting to reach your iPhone, says what went wrong when one fails, and gives you retry and delete.
• Long sends report real progress, and finished items clear themselves instead of piling up.
• Elapsed time keeps counting in always-on display rather than freezing the moment you glance at it, and the level ring dims instead of burning bright.
• The complication shows a live timer while you record, and what is waiting or failed when you are not.
• The expected speaker count turns with the crown and is remembered for next time.
• A recording can be discarded on the watch, so a pocket-tap never becomes a row in your iPhone library.
• Flag and pause have different haptics — mid-meeting they used to feel the same.

Bug fixes and performance improvements

- Faster processing
- New Apple Watch support: start recordings from your wrist, pause or resume, add markers, set an expected speaker count, and send recordings to your iPhone for private on-device transcription.
- Transcribe speech in the original spoken language, or translate non-English speech to English.
- Choose Automatic language detection or lock transcription to a specific language such as Japanese, Spanish, French, German, Chinese, and more.
- Incomplete Whisper model downloads are now detected, cleaned up, and retried instead of showing empty transcripts.
- The same transcription controls are available on iPhone and Mac.

- New Apple Watch support: start recordings from your wrist, pause or resume, add markers, set an expected speaker count, and send recordings to your iPhone for private on-device transcription.
- Transcribe speech in the original spoken language, or translate non-English speech to English.
- Choose Automatic language detection or lock transcription to a specific language such as Japanese, Spanish, French, German, Chinese, and more.
- Incomplete Whisper model downloads are now detected, cleaned up, and retried instead of showing empty transcripts.
- The same transcription controls are available on iPhone and Mac.

What's new in 1.2:

• Templated summaries — pick a format (one-on-one, standup, sales call, interview, lecture, clinical SOAP, legal, user research) for a structured, speaker-aware recap with action items and follow-ups.
• Notes mode — time-anchored notes that tap-to-seek to the moment and fuse into the summary as context.
• Labelable flags — mark and name the moments that matter while you record, then jump straight back to them.
• Recognized-speaker chips — when a saved voice is matched, the line shows a "Recognized" chip.
• On-device Apple Intelligence summaries on iPhone 15 Pro or newer (iOS 26) — no key, no network.
• Export straight into Obsidian or Apple Notes in one tap.