Google and DeepMind announce “Gemini 3.8 Live with Live Avatar,” a new version of Gemini’s conversational AI designed to add near real-time visual presence through an on-screen “live avatar.” The release is presented as an update to how Gemini interacts during live conversations, focusing on more immediate, visual-style engagement rather than text-only exchanges.
Both sources describe the same core product announcement, emphasizing that the avatar aims to be responsive in close to real time. The main difference is presentation: the Keyword frames the update as a way to bring “near real-time visual presence” to Gemini, while the DeepMind post serves as the formal technical and product introduction. Neither source, in the material provided here, specifies detailed availability, supported languages, or device platforms.
Overall, the outlets align on the central point that Gemini 3.8 Live adds a live avatar experience intended to make interactions feel more immediate and visually present, using a conversational interface built around real-time responsiveness.
