The useful question is what changes for users, developers or buyers, and whether the announcement stays industry context or becomes something people can actually use.

Writing on a phone remains cumbersome due to small keyboards, unreliable autocorrect, and distractions while moving. Voice dictation tools often produce raw transcripts filled with errors, requiring as much effort to correct as typing manually. Grammarly’s new speech-to-text feature in its mobile keyboard aims to eliminate this inefficiency by converting spoken words into ready-to-send text directly within any app. The feature is designed to handle diverse languages, accents, and speaking styles, producing polished output without the need for post-processing.

The speech-to-text function is integrated into the Grammarly Keyboard, accessible via a mic icon. Users can dictate naturally, and the system automatically removes filler words, self-corrections, and grammatical errors while preserving tone and intent. The feature supports seamless transitions between speaking and typing, allowing edits within the same keyboard interface across all apps. Grammarly’s AI assistant is also available in the keyboard for further refinements.

Grammarly’s speech-to-text is targeted at frequent mobile writers, including professionals drafting messages between meetings or students capturing ideas on the go. The feature prioritizes clarity by filtering background noise and providing visual indicators during use, such as a signal in the keyboard and the iOS orange recording light. Users retain control over when the mic is active, ensuring privacy and precision in their recordings.

The dictated text is processed into polished messages and immediately deleted after conversion, with no storage or account linkage. This ensures user data remains private and is not used for model training. The speech-to-text feature is available within the Grammarly Keyboard, which must be set as the default keyboard to function across all apps.