Bug fixes and performance updates to help you capture ideas on the go.

Speaker labels now belong to the recording, not to the app. VivaDicta 3.11.0 makes speaker labels a per-mode setting and gets you back to more of the apps you dictate from. Speaker Labels Per Mode Speaker labels used to be one switch for the whole app, so a meeting and a quick note had to agree. Each mode now carries its own setting: a Meeting mode can label who said what while a Note mode returns clean prose from the same provider, with no trip to Settings in between. Your current setting carries over to every mode, so nothing changes until you want it to. Back to 19 More Apps Dictate from the keyboard and VivaDicta returns you to 19 more apps than before. Another 16 apps that publish no way back are now recognized as such, so instead of trying and failing they take you straight to the manual return with your recording already running. Apps That Lose Your Place Safari, Messages and Claude open a new tab, a new chat or a blank new message when VivaDicta sends you back, which is not where you were typing. A new Advanced setting, Skip Apps That Lose Your Place, asks you to swipe back for just those apps instead. Swiping back returns you exactly where you were. It is off by default, so the automatic return stays the same unless you turn it on. Quality of Life Cleaner endings: the filter that removes the phantom phrases speech models produce during silence had become slightly too eager in 3.10.1 and could trim a short real phrase off the end of a transcript. The keyboard session timeout now offers 30 seconds or 1 minute, instead of five options that mostly existed for testing. Two models moved on: Groq's qwen3.6-27b rate-limited every request it was given, and Cerebras retired gemma-4-31b, so anyone using either is carried forward automatically. The top-ten model list on the AI Providers screen has been remeasured, with Gemini's gemini-3.5-flash-lite entering at rank 10.

A fix-only update. VivaDicta 3.10.1 gets the app opening again on iOS 18, clears stray marker text out of on-device transcripts, and moves the Cerebras default to a model that is still available. The App Opens Again on iOS 18 On iPhones and iPads running iOS 18, version 3.10.0 closed immediately on launch, before anything appeared on screen. It opens normally again. Nothing you had recorded was lost while it would not start, and your notes are all still there. Clean On-Device Transcripts Transcribing with Whisper on your device could wrap what you actually said in the model's own internal codes, so a short recording came back with a run of bracketed markers before and after your words. Those markers are gone. A second safeguard now removes them anywhere they turn up, including in the keyboard and in the Share and Action extensions. Cerebras Default Updated Cerebras retired Gemma 4 31B from its service, so the default model for Cerebras is now Qwen 3.8 27B, the replacement Cerebras names. If you picked Gemma 4 31B yourself it stays selected for as long as Cerebras keeps answering for it.

Nothing you said goes missing. VivaDicta 3.10.0 keeps your recording when a call cuts it short, strips the phantom lines Whisper invents out of silence, and stops opening a note from pausing your music. No More Phantom Text On-device Whisper answers a silent stretch of audio with something it memorized during training rather than with nothing, so a line you never dictated could land in your transcript. VivaDicta now reads the model's own confidence numbers to tell an invented line apart from quiet real speech, and drops it before it reaches your note. English, Russian, Turkish, Czech, and Arabic are all covered. Your Recording Survives A phone call, an alarm, or Siri used to stop the capture at the system level without telling the app, so the waveform flatlined and the minutes you had already spoken were gone. An interruption is now treated as a Stop you never pressed, and whatever was recorded up to that point is transcribed and saved like any other note. When transcription itself fails, the recording is kept as a note you can play back and retry instead of disappearing with the error. Your Music Keeps Playing Opening a note used to pause whatever you had going in Spotify, YouTube, or a podcast app, without a single tap on play. VivaDicta now takes over the audio only when you actually start playback, and hands it back when you stop or close the note. The Newest Models GPT-6 Astra, Claude Fable 5.1, Gemini 3.8 Flash, and GLM-5.3 join AI processing, and Gemini 3.8 Flash becomes the new Gemini default for both transcription and processing. Cartesia Ink-2 and Gemini 3.5 Transcribe Live join the real-time models that transcribe while you talk. Czech Keyboard A proper Czech QWERTZ layout, with y and z where Czech expects them and callouts carrying only Czech letters instead of a pan-European mix. A language added in a later update now switches itself on if you already use it in iOS, rather than waiting to be found in Settings. Back to 30 More Apps The keyboard's return list grew by thirty apps, VK Messenger among them, so dictating from one of them sends you straight back instead of showing the manual return prompt. A Calmer Settings Screen The Transcription and AI sections had grown into walls of toggles. The Gemini thinking level and transcription instruction moved down into Advanced, and the two extraction toggles now sit together behind a single Reminders and Calendar row. The Live Translation button in the toolbar is off by default on fresh installs; if you already have it, it stays. Quality of Life A recording that fails to start now tells you why, with a shortcut to Settings when the microphone was denied, instead of doing nothing visible. A downloaded on-device AI model is recognized as ready everywhere in the app, so the mode editor no longer asks you to download a model you already have.

Dictate anywhere, and land right back where you started. VivaDicta 3.9.0 sends you straight back to the app you were typing in, brings real-time dictation to the keyboard, and turns your notes into calendar events. No App Switching Dictate from the keyboard in any app and VivaDicta sends you straight back to where you were typing, instead of leaving you stranded in VivaDicta. Real-Time Dictation in the Keyboard The speed boost added in 3.8.0 now works when you dictate from the keyboard, not only inside the app. Your speech is transcribed while you talk, so the text is ready almost the moment you stop instead of waiting for the recording to upload. If the connection drops it quietly falls back to the normal upload, so nothing is lost. Speak to Edit The keyboard's rewrite tab gains a microphone. Say what you want done, like "make this more formal" or "shorten to two bullets", and it is applied to your selection, or to the text before the cursor when nothing is selected. Previously the tab could only apply a preset you had already saved, so anything outside your catalog meant leaving the app you were typing in. More Real-Time Models Streaming is no longer one provider. Deepgram Nova 3, Deepgram Flux, ElevenLabs Scribe, and Voxtral Realtime all transcribe as you speak, alongside Soniox Realtime. Also new: Gemini 3.5 Transcribe, Google's dedicated speech-to-text model with 85+ languages, speaker labels, and hesitation sounds removed as it transcribes. Your Notes Can Fill Your Calendar Reminder extraction now recognizes calendar events too. Anything that happens at a set time, like a dinner, a meeting, or an appointment, is offered as an event with its start, end, and location and goes to Apple Calendar, while to-dos keep going to Reminders. Turn it on with Extract Calendar Events Too in Settings. Transcribe With Your Grok Subscription A SuperGrok or X Premium sign-in now covers xAI Speech-to-Text as well as Grok chat, so you can transcribe without an API key. If xAI turns the subscription down for speech, it falls back to your API key when you have one. Share a Note the Moment It's Ready Turn on Auto Share Note in Settings and the share sheet opens by itself as soon as a recording finishes transcribing, so dictating a message to send someone takes one tap instead of four. Know Which Model to Pick The AI Providers screen now opens with the top ten models by measured quality and speed, each one graded on cleaning up real dictation. Models you can use without paying carry a Free tier badge. Gemini 3.7 Flash Google's newest Flash model joins both transcription and AI processing. You can also choose how hard Gemini thinks before answering and write your own transcription instruction, both in Settings. Quality of Life Notes carrying a tag you mark as excluded now survive auto-delete, keeping their audio as well as their text. A recording too long for a provider is refused up front with the limit in minutes instead of a raw error. Gemini no longer buys a reasoning pass it has no use for on every transcription, roughly halving its round trip, and xAI no longer sends your speech as English when the language is set to auto-detect. Adding an OpenAI API key no longer fails its check.

Your words, ready the moment you stop talking. VivaDicta 3.8.0 adds real-time transcription, finally records through your headphones, and refreshes the transcription and AI model lineup. Soniox Realtime A new transcription option that sends your speech to Soniox while you talk, instead of uploading the recording after you stop. The text is ready almost the moment you finish. Pick "Soniox Realtime" in the model list to use it. If the connection drops it quietly falls back to the normal upload, so nothing is lost. The standard Soniox model also returns results with noticeably less waiting. Your Headphones' Microphone If you're wearing AirPods or a Bluetooth headset, VivaDicta now records through them instead of falling back to the iPhone microphone - the same way Phone and Voice Memos behave. A new Microphone setting lets you pin the iPhone microphone instead, which keeps your headphones in full-quality audio while you dictate. On iOS 26, recording uses Apple's high-quality Bluetooth mode where supported, so music and podcasts no longer drop to muffled headset sound. Sign In With Grok Connect a SuperGrok or X Premium account and use Grok for AI processing without an API key. Sign-in happens in the browser with a short code, the same way you'd connect a TV app. xAI limits which plans can connect an app, so an API key is still there as a fallback and unlocks the full model list. Smarter Filler Removal Hesitation sounds like "uh" and "um" are now removed based on the transcript's language, so real words are no longer stripped by mistake. German "um" (as in "um zu") and Portuguese "um" stay intact. A new Remove Filler Words toggle in Settings > Advanced turns the cleanup off entirely and keeps every word exactly as spoken. New Transcription Models AssemblyAI Universal-3.5 Pro is now the default AssemblyAI model, with native code switching for multilingual speech and better speaker labels. Soniox moves to v5 for higher accuracy across 60+ languages. Also new: Gladia Solaria-3 for English, French, German, Spanish, and Italian, Speechmatics Melia-1 which detects and switches between 56+ languages on its own, OpenAI GPT-4o Transcribe Diarize for speaker-labeled transcripts, and full-size Whisper Large V3 on Groq. Latest AI Models Claude Sonnet 5, the GPT-5.6 family, Grok 4.5, and GLM-5.2 join the lineup, and Kimi and Cerebras get fresher defaults. The GPT-5.6 models are available on the OpenAI sign-in path too, so you get them without an API key. Retired models are upgraded to their replacements automatically, so no mode breaks and nothing needs re-picking. Quality of Life The keyboard session timeout gains a Never option, hot mic sessions get an in-app stop control, and adding your own transcription server now lets you choose how audio is sent, so endpoints that only accept it inside a JSON request finally work.

VivaDicta 3.7.2 is a stability update. Stability Fixed a crash that could occur when your device ran low on memory. VivaDicta now frees on-device AI models under memory pressure as intended, so the app stays smooth and reliable even during long sessions.

VivaDicta 3.7.1 brings on-device AI to the keyboard. On-Device AI in the Keyboard You can now clean up and rewrite your dictation with on-device AI right from the VivaDicta keyboard, fully offline, without opening the app. Download a small on-device AI model in AI Providers, choose it in a mode, and it runs entirely on your iPhone's Neural Engine. VivaDicta recommends Apple Foundation Model for on-device AI on devices that support Apple Intelligence; this adds an on-device option that also works from the keyboard, even on devices that do not have Apple Intelligence. Faster, More Reliable On-Device AI On-device AI now loads and runs more smoothly, starts faster, and manages memory more carefully so it stays reliable across the app and the keyboard.

Run AI on your iPhone, fully offline. VivaDicta 3.7.0 brings on-device AI text processing, plus two new cloud AI providers and clearer model options. On-Device AI VivaDicta recommends Apple Foundation Model for on-device AI on devices that support Apple Intelligence. Now, if your device doesn't - including many on iOS 18 - you can use other on-device AI models too. Download a small AI language model (LLM) like Gemma, Qwen, or Phi, and VivaDicta can clean up, rewrite, and format your dictation completely offline - no API key, no internet, nothing leaving your device. Privacy is paramount in VivaDicta, so your words never have to reach the cloud. OpenCode Zen and Go Two new cloud AI providers join the lineup. OpenCode Zen includes free models you can use with just an API key, no payment method required, and OpenCode Go adds its subscription models. Add your key once and both work. Clearer Ollama Cloud Models Ollama Cloud now shows which models are free and which need a subscription, so there are no surprises when you pick one. Performance and Polish The custom AI provider screen now scrolls fully so every field is reachable, on-device generation is more reliable, and we fixed a handful of edge cases across AI processing.

Clearer transcripts, smoother dictation. VivaDicta 3.6.0 brings speaker labels to two more cloud providers, a clearer keyboard return prompt, and one-tap copy in Live Translation. Plus performance and stability fixes. Speaker Labels for ElevenLabs and xAI Turn on Speaker Labels in your mode's transcription settings and ElevenLabs and xAI now return speaker-separated transcripts, so it's clear who said what in meetings, interviews, and group chats. Clearer Keyboard Return When the keyboard needs you to swipe back to the app to finish dictating, a new full-screen prompt makes the next step obvious, instead of a small toast that was easy to miss. Copy from Live Translation New copy buttons sit on each Live Translation column, so you can grab the original or the translated text with a single tap. Performance and Polish Filler words are now removed cleanly from translated notes, Live Translation starts faster, and we fixed a rare crash that could happen while transcribing in the background.

More ways to transcribe, process, and personalize. VivaDicta 3.5.0 adds AssemblyAI as a new cloud transcription provider, two new AI processing providers (Ollama Cloud and MiniMax), and a new Recording Orb picker so the recording screen looks the way you like. Plus performance and polish. AssemblyAI Transcription A new cloud transcription provider with high accuracy, speaker labels (diarization), and broad language support. Add your AssemblyAI API key in Settings > AI Settings > Transcription to try it. Ollama Cloud Run Ollama's hosted AI models without running your own server. Add an API key from ollama.com and pick a model in your mode's AI settings. Separate from the existing local Ollama option, and the model list refreshes automatically once your key is added. MiniMax AI A new AI processing provider for cleaning up and rewriting your transcriptions. Add your MiniMax API key in Settings and choose any of its M-series models. Recording Orb Pick the visualization shown while you record: Particles, ASCII, or None. The orb is also lighter and smoother now, so it won't bog down on busy devices. Quality of Life The app version and build now show at the bottom of Settings, the recording orb uses far less power, and a round of polish across the app.

Tag as you talk. Make VivaDicta yours. VivaDicta 3.4.0 lets you tag notes while you record, adds xAI Speech-to-Text as a new cloud transcription provider, gives Live Translation flexible save options, and brings new ways to tailor the app to how you work. Plus the latest AI models and a round of polish. Tag While You Record Add tags right from the recording sheet, so your notes land already organized. No need to open each note afterward. If you start a recording with a tag filter active on the main screen, that tag carries over automatically. xAI Speech-to-Text A new cloud transcription provider powered by Grok. Add your xAI API key in Settings > AI Settings > Transcription to try it. Live Translation Save Options When you save a Live Translation session as a note, you now choose what to keep: both languages side by side, just the original, or just the translation. Empty options are hidden so you never save a blank note. Default AI Mode Set a fallback AI mode for AI Actions. When the current note's mode has no AI processing set up, AI Actions now opens ready to use with your chosen default instead of an empty screen. You can still switch modes from the sheet. Make It Yours New Advanced toggles let you hide Chats and Live Translation if you don't use them, keeping the interface focused. Trim Trailing Period is now a single global setting instead of a per-mode toggle. Latest AI Models Added Anthropic Claude Opus 4.8, refreshed the Grok, Z.AI, Kimi, and Gemini model lineups, and removed models the providers have retired. Quality of Life Edit a note directly from the pencil button when the voice recording button is enabled, folder re-exports now overwrite the existing file instead of creating duplicates, a smoother notes list, and a livelier recording orb.

Dictation, in your language. VivaDicta 3.3.0 adds multi-language keyboard layouts, a new Cartesia cloud transcription provider, automatic folder export for your notes, and a manual Send to Obsidian button. Plus a way to try a different AI mode for a single transcription without changing your default, and small polish across Settings. Multi-Language Keyboard Type in your language. New French, German, Spanish, and Russian layouts, with a toggle key to cycle between English and your active languages. Languages you already use on the iOS keyboard are enabled automatically the first time you open VivaDicta Keyboard settings. Cartesia Transcription New cloud transcription provider with fast, high-accuracy results. Add your Cartesia API key in Settings > AI Settings > Transcription to try it. Auto Export to Selected Folder Pick a folder once and every new transcription is exported there as markdown automatically. The new Export Settings screen lets you choose the destination and toggle auto-export, and there's a one-tap manual Export to Folder action on each transcription too. Send to Obsidian Button Send any single transcription to Obsidian on demand from the transcription detail view, even when auto-append is off. Useful for picking out only the notes you actually want in your vault. Temporary Mode Override Pick a different AI mode just for one transcription right from the AI Actions sheet. The override applies once and your default mode stays unchanged. Quality of Life Auto-enable keyboard languages from iOS preferences, the floating voice-append mic on the transcription detail screen no longer hides behind the bottom action bar, refined HUD contrast for the dark theme, and small polish across Settings.

Real-time translation, two new transcription providers, and cleaner casual messages. VivaDicta 3.2 introduces Live Translation - speak in one language and hear it back in another in real time. Two new cloud transcription providers join the lineup, and Soniox, Gladia, and Speechmatics now translate as they transcribe in a single step. Plus automatic Obsidian saves, and a new toggle that strips trailing periods so casual messages don't read as cold. Live Translation Tap the new globe icon for a live interpreter. Pick a source and target language, choose whether to also hear the translation through your headphones, and adjust playback speed. 60+ languages, any-to-any, with your choice of voice. Save the session as a note when you're done. Powered by Soniox. Translate While You Transcribe With Soniox, Gladia, or Speechmatics selected, pick a target language in the new "Translate to" picker. Your transcription comes back already translated, in a single step. No extra AI processing required. Two New Cloud Transcription Providers - Gladia Solaria: 100+ languages, speaker labels, and inline translation. - Speechmatics Enhanced: industry-leading accuracy on European languages, with speaker labels and built-in translation. Trim Trailing Period New per-mode toggle that strips a trailing "." or "..." from the final transcript. Helpful for casual messages where "Okay." can read as cold. Question marks and exclamation points are kept. Cleaner Transcripts in More Languages Filler words like "uh", "um", and "hmm" are now also stripped in Russian (ээ, э-э, эм), Spanish (ehm), German (äh, ähm), and French (euh) when the mode language is set or auto-detected. Save to Obsidian Automatically Turn on Append to Obsidian in any mode and each transcription opens Obsidian with a new note. Choose one note per transcription with timestamped filenames, or pile them into a daily note. Reminders in the Original Language Extract Reminders now keeps the original language of your note. German notes produce German reminders, Russian notes produce Russian reminders, instead of occasionally being translated to English.

Notes and AI, always one tap away. VivaDicta 3.1 adds an AZERTY keyboard layout, home screen widgets, long-press Quick Actions, and new Shortcut actions so you can record, search, and chat without opening the app. Plus All Notes chat, Markdown export, and chat via OpenAI and Gemini sign-in. AZERTY Keyboard New French AZERTY layout with proper callouts and accented characters. Enable in Settings > Keyboard. Home Screen Widgets New Ask & Record and Quick Actions widgets put recording, chat, and search one tap away from your home screen. Shortcut Actions & Spotlight New Start, Stop, Cancel, and Record-and-return shortcut actions. Plus a Find Notes action and Spotlight integration so Search and Ask AI show up right in iOS search. All Notes Chat Chat across your entire library without picking individual notes. Ask broad questions and get answers grounded in everything you've recorded. Chat with AI Accounts Chat now works with OpenAI and Gemini sign-in - no API key needed. Copilot OAuth sessions stay alive across backgrounding. Markdown Export Export single notes as Markdown straight from the share menu, with a content setting to control what's included. Quick Actions Long-press the VivaDicta app icon for instant Search, Ask AI, and Need help shortcuts.

Dictate anywhere. Now talk to your notes. VivaDicta 3.0 brings Chats, Smart AI Search, Speaker Labels, and Reminder Suggestions - ask questions across your notes, find ideas by meaning, and turn recordings into action items. Chat With Your Notes Ask questions about one note or many. Surface insights, action items, and key points, or connect ideas across dozens of recordings. Chat can reach into your whole library or pull live results from the web. Powered by Apple Foundation Model on-device, or your favorite cloud AI. Smart AI Search Find notes by meaning, not just keywords. Ask "what did I say about the pricing idea?" and VivaDicta finds it - even without those exact words. Fully on-device, so your data stays private. Speaker Labels Conversations finally make sense. Get speaker-separated transcripts for meetings, interviews, and group chats, so it's clear who said what. Reminder Suggestions Turn notes into reminder suggestions, review them, and send approved items straight to Apple Reminders. Quality of Life Export notes to Markdown for Obsidian, AI assistants, or any workflow. Append follow-up recordings to existing notes without breaking your flow. Hide presets you don't use to keep pickers focused. Liquid Glass Design Enjoy a polished iOS 26 interface with Liquid Glass, plus refreshed recording controls on the keyboard and Apple Watch.

Bug fixes and performance updates to help you capture ideas on the go.

Apple Watch App Record voice notes directly on your Apple Watch. Tap the mic, speak, and your recording is automatically sent to your iPhone for transcription - even if your phone is in another room. Notes are transcribed in the background so they're ready when you open the app. Watch Modes Switch between Viva Modes right on the watch to use different transcription and AI processing settings per recording. Complications & Action Button Add VivaDicta to your watch face for one-tap recording. Assign the Action Button to start and stop with a single press. Quick Record also available in Control Center and Smart Stack. iPhone Recording Control on Watch Your iPhone recording control appears on Apple Watch under "From Your iPhone" - start and stop recording right from your wrist without the watch app.

Fixed an issue where the VivaDicta keyboard would not load on iOS 26.4. The keyboard now works reliably.

What's New in VivaDicta 2.1 Keyboard AI Processing Select text or place your cursor, then pick an AI preset — translate, rewrite, summarize, and more. Works in any app, right from the keyboard. Toggle between Voice, Text Processing, and Recent Notes with one button. Keyboard Recent Notes Quickly insert recent transcriptions from the keyboard. Your last 5 notes are just a tap away — no app switching required. Multi-level undo lets you revert consecutive pastes one at a time. AI Subscriptions Already paying for ChatGPT, Gemini, Copilot, or Claude? Sign in and use your existing subscription directly — no API keys needed. GitHub Copilot users get access to Claude, GPT, and Gemini models, including the free tier. VivAgents Server routes AI processing through CLI agents on your Mac or VPS. Cohere Transcription New cloud transcription provider with best-in-class accuracy (5.4 WER) across 14 languages. Includes a free trial tier — no credit card required. Tags & Organization Organize your transcriptions with custom tags — pick a name, color, and SF Symbol icon. Source tracking automatically labels where each note came from: In-App, Keyboard, Share/Action Extension, or Mac. Filter your history by tags or source using the new chip bar. Bulk-assign tags in selection mode. Everything syncs across devices via iCloud. New Translation Presets Eight new built-in translation presets: Chinese, French, German, Portuguese, Japanese, Korean, Arabic, and Italian. Speak in one language, get text in another. Edit Notes Edit your transcription text directly from the note detail screen. Tap the edit button in the bottom toolbar to make changes. Quality of Life • Improved mode editing with validation feedback — save button scrolls to the first error with a shake animation • Orange error badges on section headers help you spot issues quickly • Various keyboard enhancements and polish

What's New in VivaDicta 2.0 AI Presets & Variations Process your transcriptions with 40+ built-in AI presets — Summary, Action Points, Email, Coding, Assistant, and more. Each result is saved as a variation so you can compare outputs side by side. Create your own custom presets and mark favorites for quick access. AI Assistant A new Assistant preset that answers questions and transforms text. Fact-check, explain, reformat, change tone — just say what you need. Auto-Translation Speak in one language, get text in another. Dictate in English, receive output in Spanish — or any language. Set a Translate preset as your mode's default and every recording is automatically translated. Clipboard & Selected Text Context AI processing can now use your clipboard or selected text as additional context. Copy a message and dictate your reply, select a paragraph and ask the Assistant to rewrite it, or paste an email and voice-dictate your response. Configurable per mode. Keyboard Improvements • Swipe to switch between modes • Recording timer replaces the orb for a cleaner look • Improved processing status and animations VivaDicta for Mac & iCloud Sync VivaDicta is now available on macOS! Transcriptions, presets, custom dictionary, and API keys sync seamlessly across your iPhone, iPad, and Mac via iCloud. New AI Providers Added support for Z.AI and Kimi (Moonshot), plus updated model lists across all providers. Quality of Life • Auto-copy transcription to clipboard after recording • Copy button on each transcription in the list • Searchable AI model picker • Multi-select for bulk deletion • Improved Spotlight indexing • Paragraph formatting for AI-processed text • Redesigned detail view with bottom toolbar Bug Fixes & Polish • Fixed keyboard getting stuck in processing state • Improved haptic feedback across extensions • Many UI refinements throughout the app