Google has launched a new AI dictation app called “Google AI Edge Eloquent,” which operates offline by running AI models directly on users’ phones. The app transcribes speech and eliminates filler words, producing formatted text without transferring data to the cloud.
The app is currently available only on iOS, with indications that an Android version could follow. Google AI Edge Eloquent is designed to convert unrefined speech into clean text while ensuring user data remains on the device.
The app allows users to download AI models for real-time speech transcription and error correction. Additionally, it offers options to convert speech into different formats, such as “Key points,” “Formal,” “Short,” and “Long.”
An on-device mode processes all data offline, while a cloud option is available for enhanced text cleanup through Google’s Gemini service. Users can customize the app’s vocabulary by creating a custom dictionary or importing terms from Gmail to improve recognition accuracy.
Features include tracking word count and speed, as well as searching through past recordings. Keyboard integration for iOS is expected to be implemented soon, enhancing user experience. Google AI Edge Eloquent competes in the offline AI market against apps like Wispr Flow and SuperWhisper.
References to potential support for Android devices were noted in previous descriptions of the app on the Apple App Store, raising expectations for an upcoming release on that platform.
“This app puts Google directly up against apps like Wispr Flow and SuperWhisper,” a representative stated. They emphasized that the combination of offline processing and cloud-enhanced text cleanup represents a significant advancement in dictation technology.
