The Embedding Model That Understands Everything: How Gemini Embedding 2 Unified Text, Image, Video and Audio in a Single Vector Space
#1 on MTEB English by 5.09 points. 5 modalities in a single 3,072-dimensional vector. Matryoshka to truncate to 768 without retraining. Text, images, 128s video, 80s audio, 6-page PDFs. Google solved the problem that forced teams to maintain 3 models and 3 indexes.