Meta has unveiled a new WhatsApp feature that allows users to convert speech into text before a message is sent. The tool is designed, according to the company, to simplify message composition for users who prefer dictating over typing, without requiring recipients to listen to a voice note.
Identified in the Android development channel by WABetaInfo, the feature integrates a dictation mechanism directly within conversations and is expected to work across individual chats, groups, and channels. Users will have access to a new shortcut in the text input bar, through which they can select a voice recognition language and begin recording.
Once the user has spoken, WhatsApp will display the transcription as a draft, allowing the content to be reviewed and corrected before sending. The message can then be sent as text or discarded if the transcription does not match what was intended.
According to Meta, voice processing is expected to take place directly on the device. To use the feature, users will need to download the relevant language packs in advance, enabling audio-to-text conversion without transmitting data to external servers.
Local processing will also allow the tool to function without an internet connection or under limited connectivity, provided the relevant language packs are already installed on the device.
A further characteristic of the feature is that recipients will receive a standard text message with no indication that the content was originally dictated by voice, distinguishing it from conventional WhatsApp voice notes.
The feature is currently under development for Android devices and does not yet have an official general release date.
Source: Diário Económico
Original article: https://www.diarioeconomico.co.mz/2026/09/28/trends/new-ways/whatsapp-apresenta-funcao-que-transforma-audio-em-texto-antes-do-envio/
