This feature would add first-class support for Google Meet (meet.google.com) to the Kagi Translate browser extension, specifically for translating Google Meet’s live captions (speech-to-text) into a user-selected language. The key behavior is that the translated text should replace the caption text inline in the existing caption area, so the user reads the translation in the same place they normally read captions, without needing a separate panel or side-by-side view. The purpose is to provide significantly better translation quality than Google Meet’s built-in translator while keeping the experience as close as possible to native captions.
In terms of workflow, it should be minimal and non-disruptive: the user joins a Meet call, enables captions as usual, and enables Kagi Translate for the site (or has it enabled by default). From that point on, captions are automatically replaced with the translated text, updating smoothly as speech continues. The user experience improves by making live translation readable, immediate, and consistent, while avoiding extra UI clutter or manual copy/paste.
A user would use this during any Google Meet call where they rely on captions for comprehension, especially cross-language meetings. They would select the target language in Kagi Translate (the same language selection they already use on other supported sites) and then simply read the translated captions inline while people speak. Because Meet captions arrive incrementally and can be long, the extension needs streaming-aware translation: when a new speaker begins talking, it should translate whatever caption text has accumulated for that speaker immediately and replace the inline caption text. If that speaker continues talking, the extension should check every 10 seconds whether new caption text has appeared; when it has, it should re-translate the entire currently visible caption text (not only the newest fragment) and update the inline caption replacement, while doing nothing if the text hasn’t changed to avoid unnecessary requests.
This directly extends Kagi Translate’s existing “site support” model (like YouTube/Discord) into a high-value real-time scenario: Meet already provides the speech-to-text layer, and Kagi Translate would provide the higher-quality translation layer, integrated into the same place the user is already looking. Other apps implement the same idea as a native feature—Microsoft Teams offers live captions and translated captions, and Zoom offers live transcription and translated captions—where the text updates continuously as speech continues. The proposed Kagi Translate integration would mirror the same core experience (live, continuously updating translated captions), but with Kagi’s translation quality and with an inline replacement approach that keeps the UI simple and focused.
