Gemini

Google's multimodal AI model (Dec 2023) trained across text, code, audio, image, and video.