WhatsApp is testing a new feature on Android devices that allows users to dictate messages by voice and convert their speech to text before sending them, instead of recording and sending a voice message. The feature was spotted by WABetaInfo in beta version 2.26.37.8, but it clarified that it is still under development and has not yet been activated for beta testers.
Hot Topics
About half of Libyans drink contaminated water What is the scientific reason?
Mauritania's blue treasure is in danger Will the map of fish in the Atlantic be redrawn?
The Citrus Secret Everyone's Talking About: Burn More, Crave Less & Feel Great Naturally
According to WA Beta Info, WhatsApp plans to add a dedicated dictation button within the message composition bar. When used for the first time, a brief explanation of the feature will appear, prompting the user to choose a language. The application will then enter listening mode and display an indicator that it is listening to speech.
After converting the voice to text, the user can review the message before sending it, with the option to cancel the recording instead of sending it. The feature is expected to work in individual, group, and channel chats, with the other party receiving a regular text message, not the voice recording spoken by the sender.
Audio processing on the phone
The most notable aspect of the feature is the ability to convert speech to text offline. According to available information, WhatsApp relies on language packs that are pre-downloaded to the phone, so that the audio is processed locally on the device, instead of being sent to external servers to be converted to text.
This mechanism isn't entirely new to WhatsApp, as the company uses the same technology in its voice message transcription feature, which allows users to convert received voice messages into text. WhatsApp's Help Center confirms that the transcripts are generated on the user's device and that the voice messages remain end-to-end encrypted.
WhatsApp had also begun testing additional language packs for transcribing voice messages, including German, French, Italian, Japanese and additional variations of English. These packs are stored locally on the device, allowing for offline transcription.
Although the feature is described as working offline, this does not mean that WhatsApp will allow messages to be sent to recipients during network outages. Rather, the speech-to-text process can be done locally offline, while sending the text message requires an internet connection as usual.
This also differs from the current voice message transcription feature, which converts an audio recording received by the user into readable text, while the new feature converts the user's own speech into a text message before sending.
Privacy is a top priority feature
Local processing can reduce the need to send the voice recording to an external service for analysis, potentially providing an additional layer of privacy. WABetaInfo notes that using language packs on the device means the recording isn't sent to third parties before being transcribed, while WhatsApp's Help Center confirms that voice message transcription is processed on the device.
However, the feature has not yet received an official release date, and WhatsApp is still developing it for Android devices. Also, the location of the button and how it works may change before its final release, and there is no current confirmation regarding when it will reach all users or be released on other systems.
