Support

You can transcribe the first 10 minutes of any recording for free, with speaker labels. A single one-time purchase unlocks unlimited length forever — no subscription. Translation is included: the free version translates what it transcribes, and the purchase covers whole recordings.

Within those 10 minutes nothing else is held back: speaker labels and familiar voices recognized from earlier recordings, the summary and answers to your questions, the live transcript, the live note and captions while you record, translation, your saved spellings, and copying and export work as in the full version. The purchase lifts only the time limits: 10 minutes per recording and the limit on conversation mode’s free minutes. Either way, you need a compatible device and the downloaded models.

Open the unlock screen and tap “Restore purchases”. Purchases are tied to your Apple ID and restore on any device signed in with it.

Because everything runs on your device with no cloud, Vocatim downloads the speech and AI models once. After that it works fully offline. You can pick a smaller, faster package or the full one with summaries and chat.

Vocatim separates voices in a recording and can learn a person’s voice, so it labels them automatically in future recordings. You can rename or merge voices any time.

Vocatim’s interface and the transcription languages you can pick by hand cover the same 24 languages. With a Whisper model (accurate transcription) and the transcription language set to Auto, the default, Vocatim detects the spoken language and transcribes speech in about 100 languages: every language Whisper recognizes. Accuracy differs from language to language. Some features cover fewer languages: fast transcription, the live transcript while you record and translated captions use models made for fewer languages, summaries are written in one of the 24 interface languages, and translation depends on Apple’s on-device translation (see Translation below).

In Vocatim’s settings you can pick a ready-made package or choose the models yourself. Accurate transcription uses Whisper: Whisper base, Whisper large-v3 or Whisper large-v3-turbo on iPhone and iPad, and Whisper base, Whisper large-v3-turbo or Whisper large-v3 (full) on a Mac. Fast transcription uses Parakeet v3 for European languages and Nemotron for Japanese and Korean. Summaries and answers to your questions come from Qwen3.5: 2B or 4B on iPhone and iPad (4B needs a device with 8 GB of memory), and 2B, 4B or 9B on a Mac (9B needs at least 16 GB). The default, Automatic (recommended), picks the summary model that suits your device. Each model is downloaded once and runs on your device.

Yes. You can save up to 500 spellings for names and terms, for every language or for one language only, and Vocatim suggests names it finds in a transcript. A saved spelling set to Always apply corrects new transcripts automatically after recognition, and you can undo it in any recording. The corrections are made on your device; the speech model itself is not retrained.

No. There is no server. Optional iPhone–Mac sync uses your own iCloud, which only you can access.

Yes — Vocatim runs on Mac as well, with a bigger two-pane layout and larger AI models where your Mac can handle them.

Yes. Vocatim adds actions to Apple’s Shortcuts app and to Siri: transcribe an audio file, get the latest transcript, get the latest summary, and start a recording. Transcribing a file opens Vocatim and runs on your device. The transcript and summary actions need your device unlocked and pass the text on to the next step of your shortcut.

Vocatim for Mac can let an AI app on the same Mac, such as Claude Desktop, read your recordings, read-only and only while Vocatim is open. The steps are in the guide Use your Vocatim recordings in AI apps on your Mac.

The languages Apple’s on-device translation offers on your iPhone, iPad or Mac. Vocatim lists them in its Settings › Translation and marks each one as Downloaded, Available to download or Not offered by Apple. Apple’s on-device translation does not offer Czech or Slovak. Swedish, Danish and Norwegian are not available for translation in Vocatim. Translation needs iOS, iPadOS or macOS 26.4 or later.

Yes, once a language is on your device. The first time you use a language, the system downloads it, which needs an internet connection. After that, Vocatim translates without one.

No. Vocatim translates on your device with Apple’s Translation framework and sends no text anywhere. Apple says it may collect usage and performance metrics for this framework, such as which app used it and the two languages, but not the text itself.

On iPhone and iPad, tap “Download and translate” on a recording, or “Download” next to a language in Vocatim’s Settings › Translation, and confirm the system’s download window. The Settings app shows the same languages under Apps › Translate. On a Mac, confirm Apple’s download window in Vocatim, or add languages in System Settings › General › Language & Region › Translation Languages. Downloaded languages are shared with other apps and do not count toward Vocatim’s storage.

It is machine translation, made by Apple’s models on your device. Use it to follow a conversation and find what was said; names, idioms and specialist terms can come out wrong. Vocatim marks translated text as machine translation and keeps the original one tap away, so check anything important against it.

Yes. Export the translation alone or next to the original as Word, TXT, SRT or VTT, or as Markdown or JSON. The labels inside the file are in the translation’s language, and SRT and VTT keep the timing of each line.

Yes. While you record, captions show each sentence in your language a few seconds after it is spoken. Caption mode shows them in large text and can flip them for the person opposite you. Captions need a live transcription model for the language being spoken, and Vocatim does not have one for every language, for example Chinese or Norwegian. In the free version, live translation stops after 10 minutes; the recording goes on.

Two people, two languages, one screen. Put your iPhone or iPad between you and the other person: each of you reads the conversation in your own language, and their half of the screen is turned toward them. It works with two of these eight languages: Polish, English, Dutch, German, French, Spanish, Italian and Portuguese. When you end a conversation in the app after it has lasted at least half a minute, Vocatim asks whether to keep it; otherwise the conversation is not kept. A kept conversation joins your library, where you can ask it questions, get a summary and export it. Conversation mode is on iPhone and iPad, not on a Mac, and needs iOS or iPadOS 26.4 or later, with both languages downloaded; its translation is machine translation made on your device. The free version includes 20 minutes of conversation in total, not per conversation, and conversations you do not keep count too; when the minutes are used up, the app says so and the conversation ends. The one-time purchase lifts this limit.

Yes, on macOS 26.4 or later, with languages downloaded as described above. When you also record the computer’s audio, the translated captions include the other voices in an online meeting, with no bot joining it. FaceTime calls and calls handed over from your iPhone are not captured. In caption mode, Vocatim switches to a compact window that sits next to your meeting.