Google releases AI Edge Eloquent, a free offline AI dictation and text polishing app for iOS
Google has quietly released a free iOS app called Google AI Edge Eloquent, an offline-first dictation tool that transcribes speech in real time using on-device AI models, automatically removes filler words like "um" and "ah," and polishes raw dictation into professional text.[2][3] The app, powered by Google's lightweight Gemma architecture, requires no internet connection for core functions, no subscription fees, and no usage limits, positioning it as a direct challenger to paid services like Wispr Flow.[1][2][3]
Available now on the Apple App Store with minimal fanfare, Eloquent displays live transcription as users speak, complete with a waveform visualization, according to TechCrunch and The Next Web reports.[2][3] Upon pausing, the app intelligently edits out stumbles and self-corrections, then offers transformation options such as "Key points," "Formal," "Short," or "Long" to refine the output.[2][5] A toggle allows fully offline mode, keeping all audio processing on the device via Gemma-based automatic speech recognition (ASR) models, or an optional cloud mode that leverages Gemini for enhanced text cleanup.[3][6]
Privacy-focused features set it apart from cloud-reliant competitors. Users can import keywords, names, and jargon from their Gmail history or add custom vocabulary to a personal dictionary, improving accuracy without sending data off-device.[2][3][5] The app also tracks transcription history, including words-per-minute speed and total word count, as noted by Chrome Unboxed.[5] Unlike Apple's built-in free dictation, which lacks filler removal and text polishing, Eloquent delivers noticeably superior results for frequent users.[3]
This launch under the Google AI Edge brand highlights the company's push toward on-device AI, joining tools like the AI Edge Gallery app for downloading Gemma models on iOS and Android.[3][6] While currently iOS-only, the App Store description mentions "seamless Android integration" as a default keyboard with a floating button for system-wide access, similar to Wispr Flow—though Google has not confirmed a timeline.[2][5] TechBuzz and Times of India reports emphasize its role in edge computing, enabling fast, private use in low-connectivity areas.[1][4]
The move disrupts a market dominated by premium apps charging $85 to $180 annually, making advanced dictation accessible to iPhone users without barriers, as The Next Web points out.[3] Developers and enterprises may view it as a showcase for Gemma's real-world potential, while consumers gain a robust alternative to typing or basic voice tools.[3][6] For professionals dictating emails, code, or notes, this could streamline workflows significantly, especially offline.
No official announcement from Google has surfaced, and outreach for Android details remains unanswered, per TechCrunch.[2] As adoption grows, it could pressure rivals to innovate on privacy and cost, reshaping mobile productivity tools amid rising demand for local AI.[1][2] iOS users can download it immediately from the App Store to test its capabilities.