Official Google videos on Gemini — launches, keynotes, and multimodal demos.
4:35
Google's original launch video for Gemini — introducing a multimodal model built to recognize, understand, and combine text, images, audio, video, and code, available in Ultra, Pro, and Nano sizes.
2:15
Google DeepMind introduces Gemini 2.0, enabling new multimodal AI agents that can see, hear, plan, remember, and take action — powering research prototypes like Project Astra and new robotics applications.