[Digital Today reporter Jinju Hong] Google has rolled out globally a feature in Gemini for macOS that can be launched using only the keyboard's 'fn' key.

On July 29 (local time), IT outlet Engadget reported that Mac users can now press and hold the 'fn' key at the bottom left of the keyboard to talk directly to Gemini, regardless of which window they are in.

The key to the feature is cutting the number of steps to launch Gemini and attaching it directly to Mac workflows. Users can dictate immediately at the current cursor position without switching apps. Gemini is used to leave notes while looking at reading material or to record spoken content while organising thoughts verbally.

Google also added a function to refine voice transcription results. The company said Gemini automatically removes 'ums' and 'ahs' to return a more polished transcript. It aims to make the output usable immediately as a document draft or in note form, rather than simple dictation.

Users who turn on screen recognition can broaden how they use Gemini. Gemini can understand what is displayed on the screen and identify open apps. For example, if a user selects some sentences in a PDF document and then presses the 'fn' key to request a summary, Gemini produces results based on that content.

Task automation also continues. With an account that can use Spark, an agent-type AI, a user can select a sentence in notes and instruct it to draft an email based on the content. Image generation can also be carried out based on voice commands or ideas shown on screen. Google previously added Spark to its Mac app and included functions for complex tasks such as creating new spreadsheets or documents.

The Gemini app for Mac was first released in April. Google released the native macOS app the day after unveiling a native Windows 11 app, and in June it added Spark to expand the range of tasks it could perform. The change this time is that the launch feature has been expanded to 'all Gemini for macOS users.'

Language support remains limited. The feature currently supports only English, and Google said support for more languages will follow. As a result, shortcut-key launching and screen-recognition-based tasks are expected to spread first among English-language users, with expanded multilingual support likely to be the point to watch.