Google is gradually incorporating its advanced Gemini large language model (LLM) into a range of applications, including replacing older tools like Google Assistant. Among the new features is Gemini Live, a voice-based conversation mode that uses user data to tailor responses and could soon adapt more effectively to its surroundings. This mode is designed to offer a more natural and personalized interaction experience. In the latest version of the Google app for Android, version 17.58.13, Android Authority discovered a hybrid input system. This system allows users to switch between voice and keyboard input during a Gemini Live session. When activated, a swipe on the screen brings up the keyboard, enabling users to type their requests. This feature is especially helpful in noisy environments where voice recognition might not be reliable. Once a request is sent, Gemini Live displays its response across the full screen, temporarily hiding the keyboard. Users can bring the keyboard back using the same swipe gesture. If they wish to return to voice input, they can simply tap the microphone icon again. This hybrid input system offers greater flexibility and convenience, adapting to different user needs and environments. Google has not officially announced a specific release date for this feature, nor has it confirmed plans to launch it. As with many features discovered in early code, the presence of such functionality does not always guarantee its eventual public release. Nevertheless, the integration of Gemini Live and its hybrid input system highlights Google's ongoing efforts to enhance user interaction with its AI technologies.