Multimodal
AI model that handles multiple data types — text + image, text + audio, text + video.
GPT-4V, Claude 3+, Gemini see images. Whisper handles speech. Sora makes video. The modern AI train is multimodal: one model that "sees and hears" instead of one per task.