Google has introduced a new iOS application named “Google AI Edge Eloquent,” designed for advanced dictation using AI technology. This experimental tool leverages the Gemma AI models to convert spoken language into clean, structured text in real time.
The standout feature of Eloquent is its ability to comprehend the speaker’s intent rather than merely transcribing audio. The app intelligently identifies and removes hesitations, self-corrections, and filler words such as “um,” “uh,” and “well.”
Operating on an “offline-first” principle, Eloquent processes voice data directly on the user’s smartphone after the necessary models are downloaded. Users have the option to disable cloud features for enhanced privacy, ensuring that sensitive information remains on the device.
Key features of the application include:
- Instant Editing: The AI automatically corrects errors during recording.
- Quick Formatting: Users can apply presets to quickly convert text into “official,” “concise,” or “key points” formats.
- Deep Personalization: The app can import specific names and jargon from Gmail.
- Professional Analytics: Eloquent tracks speaking speed (words per minute) and maintains a searchable history of sessions.
Currently available only for iPhone, Eloquent’s description hints at an upcoming Android version. This future release is expected to allow the app to function as the default keyboard, enabling intelligent dictation across all text input fields, from messaging apps to corporate emails.
Additionally, Android users will have access to a “floating button” for immediate transcription from any screen. If testing proves successful, Google aims to integrate dictation features into the core Android system on a global scale.
Google has launched 'Google AI Edge Eloquent,' an innovative dictation app for iOS that utilizes AI to enhance speech-to-text conversion. The app offers various features for personalized and efficient transcription, with plans for an Android version that could integrate into the operating system.