Google has introduced a new free artificial intelligence-powered dictation app designed to improve spoken input by automatically refining and transcribing speech in real time. The tool, known as Google AI Edge Eloquent, works offline and does not require any subscription or usage limits.
The app is built to enhance live transcription by cleaning up spoken language as users talk. Once speech is completed, it can automatically remove filler words such as “um” and “ah”, resulting in a more polished and structured output without requiring manual editing.
Unlike many AI-based services that rely on cloud processing, the application operates entirely offline, making it more accessible in environments with limited or no internet connectivity. This approach also enhances privacy by keeping processing directly on the device.
At present, the app is available exclusively on iOS devices. However, Google has indicated plans to expand support to Android and macOS in future updates, broadening access across its wider ecosystem.
The launch reflects a growing trend in on-device artificial intelligence tools, as tech companies continue to develop solutions that prioritise speed, privacy and offline functionality while improving everyday productivity tasks.
