Google has quietly entered the AI-powered transcription space with its new offline-first dictation app, Google AI Edge Eloquent, now available on iOS. Designed to rival apps like Wispr Flow, SuperWhisper, and Willow, Eloquent allows users to dictate on their phones without relying on an internet connection once its Gemma-based automatic speech recognition (ASR) models are downloaded. The app offers real-time transcription, making it easy to capture thoughts, meetings, or interviews on the go.
Unlike traditional dictation tools that transcribe every filler word, Eloquent automatically cleans up text by filtering out “um,” “ah,” and other speech stumbles. Users can also apply formatting options such as “Key points,” “Formal,” “Short,” or “Long” to transform their transcript into polished, professional content with just a few taps. This level of AI-assisted editing is designed to bridge the gap between casual speech and ready-to-use documents.
Eloquent also offers a flexible cloud or offline mode. In local-only mode, all processing happens directly on the device, while cloud mode leverages Gemini models for advanced text refinement. The app can import names, keywords, and jargon from Gmail, and users can add custom words for personalized transcription. Additionally, it tracks session history, searchable transcripts, words-per-minute speed, and total word count, making it a powerful tool for writers, students, and professionals alike.
Currently available only on iOS, Google hints at an upcoming Android version, promising seamless integration with Android keyboards and a floating transcription button for system-wide access. This feature would allow users to transcribe text in any app, similar to existing Android transcription solutions. Google has not yet provided a release date for the Android version but indicates that broader device support is on the roadmap.
AI-powered dictation apps are becoming increasingly popular as speech-to-text technology improves. With AI Edge Eloquent, Google is testing advanced transcription capabilities that automatically polish speech into coherent, professional text. If successful, this app could signal more sophisticated, AI-driven transcription features across Google’s Android ecosystem, making it easier than ever for users to convert spoken words into high-quality written content.
source: Techcrunch