Starting your first dictation

Mac

Follow these steps to start your first dictation on Mac.

1. Click into any text field (Slack, Google Docs, Notion, etc.) so you see a flashing cursor.

2. Press and hold the Fn key (the default hotkey). When you hear a ping or see the white bars moving, start speaking.

3. Speak your message naturally at a steady pace. Flow listens while you hold the hotkey.

4. Release the hotkey when you're done. Flow pastes your formatted text into the text field. If the text appears, your first dictation is complete.

Windows

Follow these steps to start your first dictation on Windows.

1. Click into any text field (Slack, Google Docs, Notion, etc.) so you see a flashing cursor.

2. Press and hold Ctrl + Win (the default hotkey). When you hear a ping or see the white bars moving, start speaking.

3. Speak your message naturally at a steady pace. Flow listens while you hold the hotkey.

4. Release the hotkey when you're done. Flow pastes your formatted text into the text field. If the text appears, your first dictation is complete.

iOS

Follow these steps to start your first dictation on iPhone or iPad.

1. Open the Wispr Flow app and sign in if prompted.

2. Complete the onboarding tutorial if you haven't already — you'll practice dictating during the guided setup.

3. Tap into any note, document, or text field so you see the cursor.

4. Tap the microphone button. If prompted, tap Allow to grant microphone access. The keyboard is replaced by an animated waveform when Flow is listening.

5. Speak your message at a steady pace. Flow shows your words as it transcribes.

6. Tap the check button when done. Your formatted text appears in the text field.

Android

On Android, the Flow Bubble is a floating overlay that appears over all apps when you tap into a text field.

1. Open the Wispr Flow app and complete onboarding. Grant the required permissions when prompted: microphone, accessibility service, and display over other apps.

2. Tap into any text field in any app. The Flow Bubble appears on the edge of your screen.

3. Start dictating using one of two modes:

   1. Tap-to-Dictate: Tap the bubble once. The bubble expands to show Cancel (X), a waveform, and Done (checkmark) buttons. Speak your message, then tap the checkmark when finished.

   2. Hold-to-Dictate: Press and hold the bubble. Speak while holding — the bubble shows a waveform with a pulsing recording indicator. Release when done. A cancel option appears after a few seconds if you need to stop the transcription.

4. Wait briefly while Flow transcribes. Flow inserts your formatted text into the text field. If Flow can't insert the text into the field, the bubble shows a Paste button you can tap to paste from the clipboard manually.

How do I use hands-free dictation?

Mac: Press Fn + Space to start hands-free mode. Speak without holding any key. When done, press Fn + Space again to stop and paste.

Windows: Press Ctrl + Win + Space to start hands-free mode. Speak without holding any key. When done, press Ctrl + Win + Space again to stop and paste.

Can I use a mouse button instead of the keyboard hotkey?

Yes, on Mac and Windows. Open Settings → General → Shortcuts and assign a mouse button as your trigger. You can assign Middle Click, Mouse 4, Mouse 5, and Aux Button (a special entry for mice that send extra buttons, such as some Logitech models). You can also combine a keyboard key with a mouse button (e.g., Ctrl + Mouse 4). Left and right click cannot be used as triggers.

How many shortcuts can I assign per action?

Each action can have up to 4 different shortcuts assigned, and each shortcut can contain up to 3 keys total (keyboard keys and/or mouse buttons combined).

Can I start dictation from the Flow Bar?

Yes, on desktop. Click the Flow Bar in the center of your screen, wait for the ping or moving white bars, then start speaking. When you stop, Flow pastes your text into the active text field. If the Flow Bar isn't visible on a new installation, enable it in Settings.

Will my text always go into the right field?

Flow tracks which text field you clicked into when you started dictating and inserts your text there in supported apps.

What happens when I use the Action Button to dictate on iPhone?

If you assign the built-in "Quick Dictation to Clipboard" shortcut to your iPhone's Action Button (in Settings → Action Button), you can start dictating without opening any app first. When you dictate outside of a text field, your transcript is automatically copied to the clipboard — open any app and paste to use it.

A "Quick Dictation to Notes" shortcut is also available in the Shortcuts app, which saves your dictation to Flow Notes instead of the clipboard.

Why do I see a "Flow was processing your last transcript" message?

Your previous dictation is still being transcribed — or a retry you triggered from history is still processing. Flow won't start a new session until the current one finishes. Dismiss the notification, then press your shortcut again. You can also click Cancel in the notification to cancel the in-progress transcription.

What are Smart Formatting and Backtrack?

Smart Formatting automatically formats your dictation — say "first, second" to create a numbered list, or "new line" to insert a paragraph break.

Backtrack lets you change your mind mid-sentence and Flow adjusts what you just said.

What happens when I say "press enter"?

On desktop, saying "press enter" at the end of your dictation tells Flow to press the Enter key after your text is pasted. The words "press enter" are removed from the output.

The first time you use this command, a notification explains the feature (the Enter key is not pressed on that first use). Enable or disable it anytime under Settings → Experimental.

Is there a time limit for dictation?

Mac and Windows: Sessions automatically end after 20 minutes. A warning notification appears shortly before the limit. When the limit is reached, the session ends and a notification appears with a Recover text button so you can retrieve what you said.

iOS: Sessions may have a time limit enforced by the system.

Android: Sessions have a 5-minute maximum. You'll see a warning shortly before the limit. At 5 minutes, the session ends and your text is submitted.

Why does my dictation output sound different in different apps on Android?

Flow automatically applies the writing style you've chosen for each app category. For example, dictating in a work app produces more formal output than dictating in a personal messaging app. Set your preferred style for each category in Personalization Styles in the Flow app.

What error messages might I see during onboarding?

iOS: During the "try it yourself" steps, a message appears at the top of the screen if a dictation attempt doesn't succeed:

• "Oops, we didn't hear you.. Get closer to your mic and try again!" Flow picked up silence. Move closer to your microphone and try again.

• "Poor connection! Tap the mic to try again" A network issue prevented transcription. Check your internet connection, then tap the microphone button to retry.

• "Your mic is being used by another app. Close other apps and try again!" Another app has exclusive control of your microphone. Close other apps that may be using audio, then try again.

• "Enable microphone access to use Flow" Go to your iPhone's Settings, find Wispr Flow, and enable Microphone access.

• "Oops, something went wrong.. Tap the mic to try again!" An unexpected error occurred. Tap the microphone button to try again.

Mac and Windows: On the "Test your microphone" page, if Flow can't access your microphone, an error panel appears with an explanation and a tailored action button. Flow also retries automatically in the background, so the error disappears once the issue is resolved.

• Microphone access required: Flow doesn't have permission to use your microphone. Click Open Settings to go directly to your system privacy settings and grant access.

• No microphone detected: No input device is connected or available. Click Refresh devices to check again, or Open Sound Settings to verify that an input device is selected.

• Microphone is in use: Another app (such as Zoom, Microsoft Teams, Google Meet, Discord, or FaceTime) has exclusive control of your microphone. Close the conflicting app and Flow will retry automatically.

• Selected microphone unavailable: The microphone you previously chose in Flow is no longer connected. Click Choose a different microphone to pick another input device.

• Microphone error: An unexpected problem occurred. Click Restart Flow to reset the app and try again.

What microphone error notifications might I see on Mac or Windows?

If Flow can't access your microphone when you start a session, it dismisses immediately and shows one of the following notifications:

• No microphone detected: Flow couldn't find a microphone. Connect one or check that an input device is selected in Sound Settings. Click Open Sound Settings to go there directly.

• Selected microphone is unavailable: The microphone you had pinned in Flow is no longer connected. Tap Choose Microphone to select a different one.

• Microphone unavailable: Another app may be holding the mic, or the audio driver needs a reset. Tap Troubleshoot for guidance. This often clears up on its own — try dictating again first.

• Microphone error: An unexpected problem occurred with the mic. Tap Troubleshoot for guidance.

If your microphone stops picking up audio mid-session, the waveform goes flat, and shortly afterward a "Microphone is not working" notification appears with Select microphone and Contact us buttons.

Bugs fixed in recent updates

The following issues have been resolved in recent versions. If you're experiencing any of them, update Wispr Flow to the latest version and restart the app.

• First dictation after a cold start showed an error and required retrying (iOS): Right after launching the app, starting dictation from the keyboard (tap mic → Start Flow → swipe back) could fail with an error on the first one or two attempts, requiring a retry before the waveform would appear. The first dictation after opening the app now starts reliably.

• First dictation after opening the app was slow or failed to start (iOS): The very first dictation after launching Wispr Flow could fail silently or take longer than expected before beginning. Dictation now starts faster and more reliably when you first open the app.

• Tapping "Start Flow" on the keyboard didn't reliably hand off to dictation (iOS): Tapping the "Start Flow" button on the Wispr Flow keyboard to switch to the main app and begin dictating could silently fail on the first attempt. This now works reliably on the first tap.

• Keyboard got stuck after several consecutive dictations and produced duplicate text (iOS): After multiple back-to-back dictations using the keyboard's mic button, subsequent taps could cause the same text to be inserted twice. Fixed.

• Action Button, Shortcuts, and widget dictation were unreliable when toggling quickly or after a force-quit (iOS): Starting and stopping dictation in quick succession, or launching via the Action Button, Shortcuts, or a widget right after a force-quit, could fail to start. These entry points are now more reliable.

• Onboarding microphone errors showed a generic restart message (Mac/Windows): When microphone access failed on the "Test your microphone" page during onboarding, Flow showed a generic "Try restarting Flow" toast instead of a specific explanation. Flow now shows a dedicated error panel with a clear description and an action button tailored to the problem.

• Onboarding dictation failures gave no feedback (iOS): During the "try it yourself" steps in onboarding, failed dictation attempts would fail silently with no explanation. Flow now shows a message at the top of the screen explaining what went wrong and what to do next.

• Double-tap hands-free activation only worked once on the Gmail onboarding page: During the "Try It Yourself" step in onboarding, double-tapping the hotkey to start hands-free dictation could only be triggered once. You can now double-tap to activate as many times as needed on that page.

• Undo and redo buttons appeared during iOS keyboard onboarding: The undo and redo buttons were visible on the keyboard during the onboarding flow. They are now hidden throughout onboarding and reappear once setup is complete.

• Dictation showed "Listening" even when the microphone couldn't be reached (Mac/Windows): When Flow failed to access the microphone at the start of a session, it could stay in a "Listening" state and eventually show a generic "Transcript failed to load" error. Flow now dismisses the session immediately and shows a specific notification explaining the problem.

• Retrying a history item could cancel an active recording (Mac/Windows): Clicking Retry on a history item while a dictation was in progress could interrupt and discard the active recording. Retry is now disabled during active dictation, and pressing the dictation hotkey while a retry is processing shows a "Transcript currently processing" message.

• Transcript dropped when ending dictation from the Dynamic Island (iOS): Ending a dictation session by tapping the Dynamic Island could cause the transcript to be lost. The transcript is now correctly saved to Flow Notes.

• Dictated text didn't insert correctly in Gmail on Samsung and some other apps (Android): In certain apps that use a WebView-based text field (such as Gmail on Samsung devices), dictated text could fail to appear or insert incorrectly. Fixed.

• Dictation limit notifications showed the wrong time (Mac/Windows): Warning and session-ended notifications incorrectly referenced a shorter limit instead of the actual 20-minute limit. Notifications now correctly state the 20-minute maximum.

• Microphone stayed active after cancelling dictation (Android): Cancelling a dictation session could leave the microphone running in the background. Fixed.

• Submit button was clipped off-screen when switching dictation modes (Android): Switching between Hold-to-Dictate and Tap-to-Dictate could cause the Done (checkmark) button to be partially or fully hidden. The dictation bubble now fully resets between sessions.

• Keyboard appeared blank after tapping "Start Flow" (iOS): After tapping "Start Flow" to switch to a third-party app, the keyboard could appear blank or transparent until the device was rotated. Fixed.

• Background tasks interfered with active dictation sessions (Mac/Windows): Background tasks could occasionally run during an active session, causing slowdowns or interruptions. Background tasks are now automatically paused while you're dictating and resume once your session ends.

• App crashed when adding a duplicate word to the dictionary (Android): Adding a word that already existed in the dictionary could cause the app to crash. Fixed.

• App crashed when deleting a transcript on the Home Screen (Android): Deleting a transcript from the Home Screen could cause the app to crash. Fixed.

• App crashed during screen transitions (Android): Navigating between screens could occasionally cause the app to crash. Fixed.

• Words in transcriptions were sometimes misread (Android): An error in how speech recognition processed individual words could produce incorrect output. Fixed.

• Dictation activation delayed or missed during normal typing (Mac/Windows): Pressing the hotkey while typing could occasionally cause dictation to start several seconds late or not start at all. Keyboard shortcuts such as dictation activation, transform shortcuts, and paste last text were affected.

• Keyboard writing style didn't apply (iOS): Selecting a temporary writing style using the style pill on the keyboard sometimes had no effect on the dictation output.

• Hotkey took a long time to respond (Mac): Pressing the activation shortcut could cause a long delay before dictation started.

• Keyboard layout flashed (iOS): Tapping the mic button inside the Wispr Flow app briefly showed the keyboard layout before the waveform appeared.

• Dictation wouldn't start after cancelling a session: Cancelling a dictation session while it was still starting up could leave Flow in a stuck state.

• Text appeared in all lowercase (Windows): A capitalization bug during onboarding and in-app dictation caused all text to appear in lowercase.