Live captions helps everyone, including people who are deaf or hard of hearing, better understand audio by providing automatic transcription.
To make more content accessible to more people, live captions now has the ability to provide translations and will turn any audio that passes through your PC into a single English-language caption experience. When using a Copilot+ PC, live captions instantly translates any live or pre-recorded video in any app or video platform from 44 languages into English.
Live captions works across Windows 11, making it seamless to read captions while working in other apps. Captions can be provided for audio even when disconnected from the internet. You can personalize how captions are displayed, and you can include microphone audio to make in-person conversations easier.
Important: All processing of audio and generation of captions from detected voice data occurs on-device. Audio, voice data, and captions never leave your device and are not shared to the cloud or with Microsoft. Generated captions are not stored anywhere on the device or the cloud. For more information, review the Microsoft privacy statement.
Notes:
-
Live captions is available in Windows 11, version 22H2 and later. The ability to translate is available on Copilot+ PCs running Windows 11, version 24H2 and later. For more information on the new features in Windows 11, see What's new in recent Windows updates.
-
Not sure which version of Windows you have? See: Find Windows version.
In this article
Turn on live captions
To turn on live captions, do one of the following:
-
Turn on Live captions in the quick settings (to open quick settings, select the battery, network, or volume icon on the taskbar).
-
Turn on the Live captions toggle in the quick settings Accessibility flyout.
-
Press Windows logo key + Ctrl + L.
-
Select Start > All apps > Accessibility > Live captions.
-
Go to Settings > Accessibility > Captions and turn on the Live captions toggle.
When turned on the first time, live captions will ask for your consent to process voice data on your device and prompt you to download language files to be used by on-device speech recognition. If your language is not available, you can use another language during setup.
When using a Copilot+ PC, in addition to live captions, you also automatically see translation when using English (United States) as the output language. Live captions supports translating 44 languages into English.
Personalize how your captions are displayed
To choose where captions are displayed on your desktop:
-
Select the Settings button in the live captions window.
-
Select Position.
-
Choose either Above screen, Below screen, or Overlaid on screen.
-
When you choose Above screen or Below screen, the captions window will appear as docked to the top or bottom screen edges in reserved space on your desktop, and other apps will not be blocked by the captions window. When docked to the top, you might find that live captions works well while sharing video in a virtual meeting or conversation, whereas in other cases, such as viewing videos, docking to bottom might work best.
-
When you choose Floating on screen, live captions will appear in an overlay window which you can reposition as needed to avoid obscuring other apps in use.
-
To show more lines of text in the captions window, increase the window size by using the mouse, touch, or keyboard.
Make captions easier to read
-
Select the Settings button in the live captions window.
-
Select Preferences.
-
Select Caption style. The Accessibility settings for Captions opens.
-
Under Caption style, do one of the following:
-
Select a built-in style from the dropdown menu. Use the Default built-in style to have captions displayed with colors appropriate to your device’s dark or light mode setting in Settings > Personalization > Colors > Choose your mode.
-
Select the Edit button to create a custom style that works best for you.
-
Use your microphone to caption your speech
You have the option to use the microphone on your PC to caption your own speech. When this feature is on, any audio captured by your microphone will be captioned, provided that no other audio on your device is being captioned. For example, if you use live captions during an online meeting with another person, if you speak over each other, you will only see the captions for the other person.
Note: All processing of audio and generation of captions from detected voice data occurs on-device. Audio, voice data, and captions never leave your device and are not shared to the cloud or with Microsoft. Generated captions are not stored anywhere on the device or the cloud. For more information, review the Microsoft privacy statement.
To caption your own speech:
-
Select the Settings button in the live captions window.
-
Select Preferences and turn on the Include microphone audio option. When you turn on live captions, this functionality is off by default.
To check your device’s microphone configuration, see Settings > System > Sound and review the Input options.
To adjust your privacy settings for live captions’ use of your microphone, navigate to Settings > Privacy & security > Microphone > Let apps access your microphone > Let desktop apps access your microphone. For more information about microphone privacy, review Microphone privacy.
Add and use other languages
To add other languages:
-
Select the Settings button in the live captions window.
-
Select Change language, select the desired language from the dropdown, then select Continue.
-
If the language is not already downloaded, you will be asked to download it. Select Download to confirm.
-
After the download completes, live captions will display that it’s ready to caption in the new language.
Tip: Languages already installed on your device are highlighted in bold in the language selection dropdown.
To add other languages:
-
Select the Settings button in the live captions window.
-
Select Caption language.
-
Select Add a language. The Language & region settings window opens.
-
In Language & region settings, go to Preferred languages, and then select Add a language.
-
In Choose a language to install, browse or search for a language with support for Speech recognition, and then select Next.
-
In Install language features, select the features you want to use, while ensuring Enhanced speech recognition is selected, and then select Install.
When the installation of the enhanced speech recognition feature has completed for the language you selected, the language appears in live captions' Caption language menu.
To use other languages:
-
Select the Settings button in the live captions window.
-
Select Caption language.
-
Select the language you want to use.
When the new language is selected, live captions will display that it’s ready to caption in the new language.
Get the most out of live captions
To help you understand controls you have, here are additional ways you can get the most optimal experience:
-
To mask profanity, go to the Settings menu, select Preferences, and turn on the Filter profanity option.
-
To improve captioning accuracy when using the microphone, make sure to minimize background noise in your environment and speak directly into the microphone.
-
To ensure minimal delay in captions or if you notice that captions are not appearing, try closing unused apps to maximize performance.
-
Resource-intensive apps (for example, apps that share video) might impact the real-time behavior of live captions, leading to delays in captions, or even dropped captions. If this happens, consider limiting some app functionality while depending on live captions (for example, turn off any background effects or other special effects applied to shared video).
-
Microsoft’s commitment to responsible AI
Live captions is built responsibly keeping your privacy in mind. It keeps the language files and data on the device, keeps the microphone off by default, and provides an optional profanity filter to mask profane speech elements. In addition to this, live captions with translation extends the capabilities of live captions to break language and accent barriers.
Live captions uses Azure AI Speech models, a compact version of the captions language files that are evaluated on the same fairness datasets as the cloud-based Speech to Text API. These models are embedded on the device to provide streamlined local captioning and translation with reasonable and acceptable accuracy in real-time. To get more information about responsible use of Azure AI Speech, see Speech-to-text fairness information and Transparency Note and use cases for speech to text.
For more about our responsible AI efforts, the principles that guide us, and the tooling and capabilities we've created to assure that we develop AI technology responsibly, see Responsible AI.
We want to hear from you!
If there's something you like, and especially if there's something you don't like, about live captions you can submit feedback via Feedback Hub (press Windows logo key + F while live captions is active) and select Accessibility > Live captions category.
Frequently asked questions
Live captions supports speech recognition in:
-
Chinese (Simplified, China)
-
Chinese (Traditional, Hong Kong SAR)
-
Chinese (Traditional, Taiwan)
-
Danish
-
English (Australia)
-
English (Canada)
-
English (India)
-
English (Ireland)
-
English (New Zealand)
-
English (United Kingdom)
-
English (United States)
-
French (Canada)
-
French (France)
-
German (Germany)
-
Italian (Italy)
-
Japanese
-
Korean
-
Portuguese (Brazil)
-
Portuguese (Portugal)
-
Spanish (Mexico)
-
Spanish (Spain)
On a Copilot+ PC, live captions has the ability to translate to English from these languages:
-
Arabic
-
Basque
-
Bosnian
-
Bulgarian
-
Chinese (Cantonese)
-
Chinese (Mandarin)
-
Czech
-
Danish
-
Dutch
-
English
-
Estonian
-
Finnish
-
French
-
Galician
-
German
-
Greek
-
Hindi
-
Hungarian
-
Indonesian
-
Irish
-
Italian
-
Japanese
-
Korean
-
Latvian
-
Lithuanian
-
Macedonian
-
Maltese
-
Norwegian
-
Pashto
-
Polish
-
Portuguese
-
Romanian
-
Russian
-
Slovak
-
Serbian
-
Slovenian
-
Somali
-
Spanish
-
Swedish
-
Thai
-
Turkish
-
Ukrainian
-
Vietnamese
-
Welsh
Only speech detected in audio will be captioned. Audible events such as applause or music will not be detected. Lyrics sung in music will not be reliably detected.
All processing of audio and generation of captions from detected voice data occurs on-device. Audio, voice data, and captions never leave your device and are not shared to the cloud or with Microsoft. Generated captions are not stored anywhere on the device or the cloud. For more information, review the Microsoft privacy statement.
Live captions pays attention to the default sound output device configured in Settings > System > Sound. You might need to change your default device for audio to be picked up by live captions.
The microphone is always turned off by default when live captions starts up, so that only the audio you intend will be captioned.
Sound audio will be prioritized over microphone audio. For example, if you are in a virtual meeting where a remote meeting participant is speaking and you speak over each other, the captions for the remote meeting participant will be shown instead of your own.
Navigate to Settings > Apps > Installed apps, and then search for Speech Pack. You get a list of all installed language files. Select Uninstall from the More menu for the language file you want to uninstall.
