Frequently Asked Questions

VoiceVibe is an AI voice assistant you can use anywhere on your computer. Press a key and talk. It types, speaks and does what you say.

• • 10 specialized voice modes for dictation, text editing, AI image generation, text-to-speech, and conversational AI
• • Works across any application on your desktop
• • Voice Typing is free
• • Pro subscription ($14.99/mo) unlocks all Pro modes and features

VoiceVibe supports:

• • Windows: Windows 10 or later
• • macOS: Apple Silicon Mac (M1 or newer) running macOS 14 or later
• • Linux: X11 on x86-64, with initial QA on Linux Mint

We provide native integrations for each platform with proper permissions and system-specific optimizations.

Free tier: NO!

• • Voice Typing works 100% offline with on-device Parakeet AI

Pro tier: YES

• • Requires internet connection for cloud-based AI processing
• • Voice Typing can work offline when Local Model or Cloud with Local Fallback is selected; Strict Cloud requires an internet connection

Download sizes vary by platform and release. The download page lists the current size of each installer.

Offline Voice Typing downloads the on-device Parakeet AI model on first use. Plan for about 2.9 GB on disk with the local model installed, or about 320 MB for the app with the model unloaded.

Yes! VoiceVibe includes a free Pro Preview.

The Pro Preview lets you try all Pro modes with a small usage allowance — enough to try multiple modes and get a real feel for Pro. Look for the "Try Pro" option in the app.

• • All Pro modes included except AI Images
• • No credit card required
• • Preview is usage-based, not a timed trial. An activated Preview with no usage may expire after 30 days; positive usage prevents inactivity expiry. Only an explicitly authorized reactivation resets the unused clock.
• • One Pro Preview per eligible user

No! Neither tier requires API keys.

• • Free tier: No account or API key required. First offline use downloads the local speech model.
• • Pro tier: We use a mix of self-hosted and cloud AI services, selected for speed and quality. All routing is automatic - you never need API keys or technical setup.

Yes! Each mode has its own configurable keyboard shortcut.

Choose a primary key:

• • Modifier keys: Ctrl (L/R), Alt (L/R), Shift (L/R), plus Cmd (L/R) on macOS
• • Function keys: F6, F7, F8, F9, F10, F11, F12

Optionally add a secondary key for combinations:

• • Letters (A-Z), numbers (1-9), Space, or F1-F5
• • Example: Ctrl + Space or Alt + A

Configure keyboard shortcuts in each mode's settings tab. The default for Voice Typing is Right Ctrl.

Free tier:

• 1. Voice Typing - Speech-to-text dictation

Pro tier:

• 2. Magic Typing - Create or transform text in any app
• 3. Magic Clipboard - Create or transform text to your clipboard
• 4. Voice Assistant - Conversational AI with spoken responses
• 5. Text-to-Speech - Convert any selected text to natural audio
• 6. AI Images - Generate images from voice descriptions
• 7. Agent Launcher - Dictate, review, and launch tasks for supported coding agents
• 8. Chat Launcher - Send voice prompts to ChatGPT, Claude, Gemini, etc.
• 9. Note Taker - Voice-to-structured notes
• 10. Add to Dictionary - Voice-powered vocabulary building

Magic Typing types the result straight into your app. Select text and speak an instruction (e.g., "make this more formal") to replace it with the AI-transformed version, or speak with nothing selected to create something new where your cursor is.

Magic Clipboard puts the result on your clipboard ready to paste, instead of typing it into your app. Speak an instruction to create something new, or add selected text, clipboard content, screenshots, or files as context - nothing on screen is replaced.

Chat Launcher lets you start a conversation with ChatGPT, Claude, Gemini, Perplexity, or a custom service using your voice. Speak your prompt, add selected text, clipboard content, screenshots, or supported files, and VoiceVibe opens your chosen browser or desktop app with everything ready for you to review and send.

Agent Launcher lets you dictate, review, and launch tasks for Claude Code, Codex CLI, Cursor, OpenCode, or Pi. You can review the prompt, working directory, context, and available launch permission before starting the agent. Availability varies by platform.

Agent Launcher can include selected text, Clipboard Content, screenshots, files, and pasted text. You can review and remove context before launch, and VoiceVibe hands it locally to the selected coding agent. Optional notifications can tell you when the agent finishes or needs input. A spoken summary is played only after you open the notification.

Yes! Voice Assistant supports multi-turn conversations.

By default, conversations are saved on your device so you can return to them later. You can also turn off saving for session-only conversations. Ask a question, hear the answer, and continue the conversation naturally.

Saved conversations are stored on your device, not as conversation history on VoiceVibe's servers. You can delete individual conversations, clear your history, or turn off saving for new conversations.

Yes. When Search the web when needed is enabled and available, Voice Assistant can retrieve current information for questions that require it. It does not search every time. VoiceVibe sends a generated search query - not your full conversation or attachments - to the search provider, then produces the final answer. Source links can be included in the Markdown conversation export.

Push-to-Talk: Hold the keyboard shortcut while speaking, release to stop. Best for short commands and quick dictation.

Toggle Recording: Press once to start recording, press again to stop. Best for longer dictation sessions where holding a key would be uncomfortable.

You can set the activation mode independently for each voice mode in its settings tab.

Free gives you basic speech-to-text transcription with the Parakeet model running on your device.

Pro uses cloud AI with post-processing to dramatically improve quality:

• • Auto-corrects transcription errors - lower word error rate
• • Removes filler words - um, uh, ah, like, you know
• • Cleans up self-corrections and repetition - concise output
• • Better formatting - improved punctuation and structure
• • Personal Dictionary integration - improved accuracy with your names and terms

Voice Typing can work offline when Local Model or Cloud with Local Fallback is selected. Strict Cloud mode requires an internet connection.

Pro modes support multiple input options:

• • Voice: Always available - speak your request
• • Selected text: Highlight text and include it as context
• • Clipboard: Attach clipboard content to your request
• • Screenshot (selection): Select a screen region to capture
• • Screenshot (full screen): Capture your entire screen
• • File attachments: Attach files for context

Enable input options from each mode's settings. Availability depends on the selected mode.

Select any text on your screen, press the Text-to-Speech keyboard shortcut, and hear it read aloud in a natural-sounding voice.

• • 120 preset voices plus voice cloning
• • Choose your preferred voice and speed in Speech Settings
• • Works with any selected text in any application
• • Pro subscription required

Yes! Pro users can create up to 3 custom voice clones.

Record a 5-10 second sample of clear speech, and VoiceVibe will create a voice clone you can use for Voice Assistant responses and Text-to-Speech.

Tips for best quality:

• • Record in a quiet room
• • Use a USB microphone if possible
• • Speak naturally at your normal pace
• • Provide 8-10 seconds of clean speech

Manage voice clones from Speech Settings in the app. Pro subscription required.

Free on-device Voice Typing supports 25 languages. Pro users can configure from 100 cloud speech languages and choose a default plus up to two additional languages. Auto chooses among the languages you have configured. Text-to-Speech includes 120 preset voices across supported languages, plus optional voice cloning. Availability and quality vary by language and feature.

Configure languages in Language Settings and voices and speaking speed in Speech Settings.

Free Voice Typing: YES!

• • Works 100% offline with on-device Parakeet
• • Perfect for privacy and offline work

Pro modes: NO

• • Require internet connection for cloud AI processing
• • Voice Typing can work offline when Local Model or Cloud with Local Fallback is selected; Strict Cloud requires an internet connection

The Personal Dictionary lets you teach VoiceVibe your unique vocabulary - names, technical terms, acronyms, and jargon. For each word, you can add a description that helps the AI understand the context.

Key features:

• • Works across all Pro modes for consistent accuracy
• • Add unlimited custom words and phrases
• • Include descriptions to provide context
• • Import/Export for backup and sharing
• • Stored locally on each device

Word Replacements: Simple text substitutions that happen locally on your device (e.g., "michael" → "Mikael", "pie torch" → "PyTorch").

Personal Dictionary: More intelligent - it sends your vocabulary to our AI, which uses the descriptions you provide to understand context and recognize terms correctly.

Dictionary works best for names, technical terms, and domain-specific vocabulary.

Include information about yourself and any preferences for how VoiceVibe should write:

• • Your profession, workplace, and role
• • Technical terms or jargon from your field
• • Projects or topics you often discuss
• • Personal interests and hobbies you talk about
• • Spelling and formatting preferences (e.g. "Use British English spelling")

Example: "I'm a marketing manager based in London. Use British English spelling. My team includes Priya and Dr. Okonkwo. We use HubSpot and Notion."

Yes! VoiceVibe includes a Word Replacements feature.

Use it to:

• • Fix commonly misheard words or names
• • Expand abbreviations automatically
• • Correct technical terminology
• • Personalize transcription output

Access it from the Word Replacements tab in the app. Supports case-sensitive matching.

Yes. VoiceVibe works with essentially any desktop app or website—type directly where supported, transform selected text, hear text read aloud, use screenshots for context, or use Voice Assistant to ask questions and get guidance in any application.

VoiceVibe checks for updates automatically. When an update is available, you'll see a notification. Updates are downloaded and installed manually - the app does not auto-update without your permission.

You can check for updates manually or configure update behavior in Settings.

Voice Typing mode only:

• • Basic speech-to-text with on-device Parakeet AI
• • Unlimited usage
• • Works 100% offline
• • Uses ~2.7GB RAM while the local model is loaded
• • Voice audio and transcription stay on your device
• • No account or API key required; first offline use downloads the local speech model
• • Free on a single device

All 10 voice modes included:

• • All Pro modes (Magic Typing, Voice Assistant, AI Images, Text-to-Speech, Agent Launcher, and more)
• • Faster cloud AI processing
• • Voice cloning (up to 3 custom voices)
• • Multi-device support (up to 3 devices)
• • Voice clones sync across devices; Dictionary and About You stay local per device
• • Priority support from your account dashboard
• • All-inclusive - no API keys needed from you!

Pro and Pro Max include the same 10 voice modes and features. The only difference is usage limits - Pro Max gives you 4x the usage capacity, ideal for power users who use VoiceVibe heavily throughout the day. Pro is $14.99/mo ($11.99/mo annual), Pro Max is $29.99/mo ($23.99/mo annual).

The Pro Preview gives you a taste of Pro with zero commitment:

• • All Pro modes except AI Images
• • A small usage allowance — enough to try multiple modes
• • Preview is usage-based, not a timed trial. An activated Preview with no usage may expire after 30 days; positive usage prevents inactivity expiry. Only an explicitly authorized reactivation resets the unused clock.
• • No credit card required
• • One Pro Preview per eligible user

Activate the Pro Preview from the "Try Pro" option in the desktop app.

Pro includes generous usage limits that reset monthly on both monthly and annual subscriptions. Annual billing changes when you are charged and the effective monthly price; it does not change how often your usage allowance resets. The vast majority of users never hit the limit with normal daily use. If you need significantly more capacity, Pro Max offers 4x the usage limits.

AI Images has its own fair-use allowance, resets on the same monthly date, and also counts toward your monthly usage. If the image allowance is used, your other Pro features keep working until your main usage limit is reached.

Free Voice Typing is unlimited - no usage caps, ever.

Pro modes will stop working until your next monthly usage reset. Voice Typing (on-device) continues working always - it has no usage limit.

Consider upgrading to Pro Max (4x limits) if you find yourself regularly reaching your limit.

Yes! Cancel anytime from your dashboard.

• • No questions asked
• • No cancellation fees
• • You'll retain Pro access until your current billing period ends
• • Voice Typing is included in the free tier
• • Resubscribe anytime to restore Pro access

Up to 3 devices per account

Use any mix of Windows, macOS, and Linux. Voice clones sync across your connected devices via your account. Personal Dictionary and About You context are stored locally on each device and need to be set up separately.

Manage your connected devices from your dashboard.

7-day money-back guarantee on all Pro subscriptions.

• • Not satisfied? Submit a dashboard support request within 7 days for a full refund; full refunds end paid access immediately
• • After 7 days, refunds for technical issues are considered case-by-case within 14 days of the charge
• • Cancel anytime from your dashboard

Yes! Save 20% with annual billing.

• • Pro monthly: $14.99/month
• • Pro annual: $143.90/year (equivalent to $11.99/month)
• • Pro Max monthly: $29.99/month
• • Pro Max annual: $287.90/year (equivalent to $23.99/month)

Choose your billing cycle when subscribing from your dashboard.

We process payments securely through Stripe:

• • Credit/debit cards (Visa, Mastercard, American Express, Discover)
• • Apple Pay and Google Pay (where available)

Your payment information is handled directly by Stripe and never stored on our servers.

Steps to upgrade:

1. 1. Sign in to your dashboard at voicevibe.ai
2. 2. Click "Upgrade to Pro" or "Subscribe"
3. 3. Complete the subscription signup via Stripe
4. 4. Pro access activates immediately
5. 5. In the desktop app, use the "Check for upgrade" button to verify your Pro status and unlock all Pro modes

Free tier:

• • Voice audio and transcription stay on your device
• • Processed entirely on-device with Parakeet

Pro tier:

• • Audio temporarily sent to VoiceVibe backend then to AI services for processing
• • VoiceVibe does not save voice recordings after processing
• • Voice Assistant saves your conversations on your own device until you delete them, and you can turn saving off. VoiceVibe keeps no copy on its servers
• • Files and screenshots may be temporarily cached to support follow-up questions; see the Privacy Policy for retention details

Only data necessary for processing:

• • Audio recordings for transcription
• • Selected text for editing commands
• • Limited text around your cursor in Pro Voice Typing and Magic Typing
• • Relevant Voice Assistant context when you ask a follow-up question
• • Screenshots when screenshot modes are enabled
• • File attachments when you explicitly include them

VoiceVibe does not transmit browsing history, unrelated system data, or files/screenshots you did not explicitly include. Account, device, billing, and usage metadata are handled as described in the Privacy Policy.

VoiceVibe does not currently use your content to train AI models, and we use AI providers under terms and settings that prevent customer inputs and outputs from being used for provider model training.

• • Your voice recordings, transcriptions, and text are not stored as model-improvement history
• • Voice Assistant saves your conversations on your own device until you delete them, and you can turn saving off. VoiceVibe keeps no copy on its servers; files and screenshots may be temporarily cached to support follow-up questions
• • See the Privacy Policy for current provider and retention details

Read our full Privacy Policy for complete details.

Free tier: VoiceVibe collects a pseudonymous device identifier and daily usage count for product notifications and improvements. You can disable crash reporting in Settings > Privacy. No voice audio or dictated text is collected.

This telemetry is not currently opt-out, but it contains no voice audio, dictated text, or personal content — only a random device identifier and a daily counter.

Pro tier: We log per-request metadata (mode, audio duration, request cost, timestamp) only for quota enforcement and billing. We do not run product analytics or behavioural profiling on this data. Request content is not included in these usage records.

Free tier (on-device):

• • Parakeet Ultra - an improved version of NVIDIA's Parakeet speech recognition model, by Moondream

Pro tier (cloud):

We use a mix of self-hosted and cloud AI services, selected for speed and quality. All routing is automatic - you never need API keys or technical setup.

Yes, at multiple levels:

• • In transit: All data sent to our servers uses TLS/SSL encryption
• • Authentication: Authentication tokens use OS-provided secure storage where available

Free tier users benefit from local voice processing: voice audio and transcription stay on your device.

Yes, your data is private and secure.

• • Only used during processing to improve accuracy
• • Not shared, sold, or used for any other purpose
• • Transmitted securely to our servers
• • Not stored beyond what's needed for the session

Common causes and solutions:

• • Keyboard shortcut conflict: Another app is using the same key. Try a different keyboard shortcut combination in settings.
• • macOS permissions: Grant Accessibility permission in System Settings > Privacy & Security > Accessibility.
• • Linux Wayland: VoiceVibe requires X11. Switch to an X11 session at the login screen.
• • VoiceVibe not running: Check your system tray (Windows/Linux) or menu bar (macOS) for the VoiceVibe icon.

A dedicated USB microphone gives the best results. Here's a quick comparison:

• • USB microphone: Best quality - clear audio, minimal noise
• • Bluetooth headset: Good - convenient and works well
• • Built-in laptop mic: Acceptable - may pick up keyboard and fan noise

Position the microphone about 6-12 inches from your mouth. In your OS sound settings, check that the input meter moves clearly when you speak but does not constantly hit the maximum level. If audio sounds distorted, lower the input volume. If VoiceVibe misses words or hears very little, raise the input volume or move closer.

VoiceVibe uses your system's default audio input device. Try these steps:

• • Check that your microphone is set as the default input device in your OS sound settings
• • macOS: Grant Microphone permission in System Settings > Privacy & Security > Microphone
• • Linux: Verify PulseAudio or ALSA is running
• • Close other apps that may be using the microphone
• • After changing the default microphone or input volume, restart VoiceVibe and try a short test recording

Pro modes require an active internet connection. If they're not working:

• • Check your internet connection - try loading a web page
• • Corporate firewalls may block the connection VoiceVibe uses - check with your IT team
• • VPN interference - try disconnecting your VPN temporarily
• • Restart VoiceVibe to re-establish the connection

Voice Typing can work offline when Local Model or Cloud with Local Fallback is selected. Strict Cloud mode and other Pro modes require internet.

Steps:

1. 1. Go to System Settings > Privacy & Security > Microphone
2. 2. Go to Accessibility and enable VoiceVibe (required for keyboard shortcuts)
3. 3. Go to Screen Recording and enable VoiceVibe (required for screenshot features)
4. 4. Grant Automation permission when prompted (required for focus management and text insertion)
5. 5. You may need to restart VoiceVibe after granting permissions

RAM usage depends on your tier:

• • Free tier: ~2.7GB RAM while the local model is loaded - This is normal! The on-device Parakeet AI model runs locally for privacy.
• • Pro tier: ~320MB RAM with the local model unloaded - Lower memory usage since AI processing happens in the cloud.

If you're experiencing high RAM usage on Free tier but want lower memory consumption, consider upgrading to Pro - the cloud AI processing means VoiceVibe uses significantly less local resources.

Try these steps:

1. 1. Click the "Check for upgrade" button in the app's settings
2. 2. Ensure you're signed in with the same account you used to subscribe
3. 3. Check your internet connection
4. 4. Try signing out and signing back in

If issues persist, verify your subscription status in your web dashboard. Active Pro and Pro Max users can request priority support from the account dashboard.

Not currently. VoiceVibe supports X11 on x86-64 Linux. Wayland is not currently supported.

Linux Mint is the initial QA target.

VoiceVibe supports X11 on x86-64 Linux. Linux Mint is the initial QA target.

• • .deb package: Best fit for Linux Mint and similar apt-based X11 systems
• • AppImage: Portable package for x86-64 X11 systems

To report a bug effectively:

• • Use our feedback form
• • Include system information (OS, version, hardware)
• • Describe reproduction steps
• • Attach relevant screenshots

Bug reports and feedback are welcome here; Pro and Pro Max support requests are handled from the account dashboard.

Log file locations:

• • Linux: ~/.local/share/voicevibe/logs/
• • macOS: ~/Library/Logs/VoiceVibe/
• • Windows: %LOCALAPPDATA%\VoiceVibe\Logs\

Log files are created automatically and contain details needed to diagnose issues. Include them when reporting bugs or when opening a dashboard support request.