What's new in 1.7: translation, on your device. • Translate a recording into your language. Read the original, the translation or both, line by line, and switch between them at any time. • A recording that mixes languages is translated line by line, each from its own language. When you correct the transcript, the translation follows. • Translate the summary and the live note in one tap; their sections and tasks stay in place. • Export the translation on its own or next to the original as Word, TXT, SRT or VTT. Any recording can now also be exported as Markdown or JSON. • Library search also looks inside translations, so a search in your language finds a recording made in another, once it is translated. • Translated captions while you record: each sentence appears in your language a few seconds after it is spoken. • Caption mode: the captions in large text across the whole screen, with a button that flips them for the person opposite you. • Apple Watch: pause, stop or bookmark a recording from your wrist (iOS 18 and watchOS 11 or later). Translation needs iOS or iPadOS 26.4 or later. The first time you use a language, the system downloads it once; after that, translation works offline. Translations are machine translations. Plus background playback and Lock Screen controls.

What's new in 1.6: • A new look and dark mode: Settings > Appearance: System, Light or Dark. On iOS and iPadOS 26 the recording screen uses Liquid Glass: switch sections with a glass capsule, speakers get glass buttons of their own, and the text scrolls beneath them through a soft blur. • Questions about a recording get their own view next to the summary and no longer get lost beneath it. Answers appear word by word, you can stop them, and each one shows the places in the transcript it is based on. • “Ask your meetings” answers in the language of your question and names meetings by their titles. An answer keeps coming when you leave the conversation: “Ask your meetings” shows it is still working, and a new answer is marked in the list of conversations. • See what's new at a glance: a recording's card in the library says “New transcript”, “New summary” or “New answer”, and the badge on the app icon counts those recordings. • Stop a recording that has live text and it opens right away; the more accurate version replaces the text in place when it is ready. With a paired Mac you still get the sheet that lets you choose. • Faster summaries: shorter meetings are summarized in a single pass instead of several, and decisions and tasks (who, by when) come out more reliably. • Names and spelling (the Abc button) now reviews the whole recording: search names, Apply all, and Undo after every change, with fewer and better suggestions. With Always apply, new recordings correct themselves. • You can record while models download: transcription starts by itself once they are ready, and questions and summaries show the download's progress. • Native menus and sheets: ⋯ and a long press on a card or a line open system menus (a line's menu now has Play this line and Change speaker), and sheets such as Trim close with a swipe down. Plus: library search now covers summaries and notes, times and dates follow your region's format, and many smaller things work more reliably.

On iPhones running iOS 26.1 or later, transcription and speaker recognition now finish even after you lock the screen — the system shows the progress, and Vocatim no longer has to stay in front. Start processing while the app is on screen, then put the phone away. • If the system interrupts work in the background, you get a notification and the session shows on the list that speaker recognition is waiting to be finished. The transcript stays intact; one tap finishes the rest. • Model downloads can continue while the screen is locked. Once the files are in, a notification reminds you to come back to Vocatim and finish preparing the model. • Summaries, chat and answers use the graphics accelerator — keep Vocatim on screen while they are being generated. Plus stability fixes around interruptions and resuming work.

Two ways to fix a recording after the fact — and the text you keep stays as it is. • Trim recording removes the end of a recording without cutting a transcript line in half. Before you confirm it counts what goes: removed audio, transcript lines, bookmarks, and people who appear only in the removed part. A full copy goes to Recently Deleted for 7 days, and you can have a new summary written afterwards or keep the one you have. • "Split into more than 4 people" (or back into at most 4) redoes only the speaker split. The text and your edits stay, names are carried over where the voices line up, and you can restore the previous split until you change the transcript. Dragging the playback position works again on iPad, and jumping deep into a long transcript now lands where it says.

In this update: - Edit summaries and note text directly inside a session. - Correct name spellings across the transcript, summary and notes in one go. - Save corrections for future recordings, for all languages or just one. - Sync your dictionary between iPhone and Mac with iCloud. We've also improved editing, saving changes and app stability.

New in 1.5: • "More than 4 speakers" now runs an offline engine with no speaker limit, and on a crowded meeting it separates people far better than what ran there before. The recording's Info tab says which engine made the split and how many speakers it found. • One record button on the home screen instead of two. The library gets its room back, and a long press on the button still opens the record screen. • The walkthrough on the first run points at controls that are really on the screen, and it never asks for a permission — each permission stays on the switch that needs it, where iOS asks in its own words. • The record screen says what Stop will need and where those models arrive, instead of showing you the same elapsed time twice. • The Notes tab is one system of headings, marks and colours instead of three, and the legend for the marks is there before you have asked a single question. • The model packages on the first run stop looking like an option already chosen and waiting for a Confirm: each one gives its name, its size and what it gets you, and any of the three starts on the spot. • A voice card trades twelve buttons for one menu, empty screens share one shape, and a switch is Vocatim green everywhere. • Counts agree with their words at last — one recording, two recordings — in each of the eighteen languages, by that language's own grammar. • Two names for two things, throughout: the live transcript and the accurate transcript.

New in 1.4: • The summary is written while the meeting runs. Points, decisions and tasks build up as people talk, so when you stop it is already there — with the summary model on your device. • Ask a question without interrupting. Type it during the meeting and Vocatim marks the passages that answer it; with the summary model it writes a short answer too. Question and answer stay with the recording. • A live note. Decisions, tasks, questions and figures are picked out of the transcript as it is written, together with the moments you pinned yourself. • The live preview keeps up with the conversation and shows who is talking. A voice you have saved gets its name mid-meeting. • The live preview now speaks Japanese, Korean and Turkish too. • When you stop, you decide: keep the transcript the meeting already produced, transcribe the recording properly on this device, or hand it to your paired Mac. • A recording opens in tabs: transcript, summary, info, and notes when there are any. From the summary, a note or a bookmark you jump straight to that place in the transcript — and in a wide window on Mac and iPad the transcript reads beside a panel of the people in it. • "Recording only" — one switch stops everything that would otherwise run during a take. • Export to Word: every passage labelled with its speaker and its time. • Better speaker separation in fast mode, summaries on a new model family matched to your device, and playback and chat that keep going when you leave the screen.

New in 1.3: • Vocatim now also speaks Japanese, Korean, Swedish, Chinese (Traditional), Turkish, Norwegian, Danish, Czech, Slovak and Ukrainian — eighteen languages in total. • Faster model downloads — from Vocatim's own server, in several parallel streams, resuming automatically after a dropped connection. • You can see what's going on: "Transcription model 1/2", time remaining, and real progress while the model is prepared (step 3/7) instead of a spinner. • A model package finishes even if one model fails to download — you get the rest, and the error shows once, at the end. • Have a promo code? Redeem it right from the purchase screen ("I have a code"). • A device storage bar where you choose models: how much Vocatim uses, how much other data does, how much is left. • Speaker count next to the speaker strip, and an honest, faster Mac ↔ iPhone connection status. • VoiceOver reads download progress aloud, plus stability fixes.