Google DeepMind Introduces Multimodal Embedding Model for On-Device AI

Google DeepMind has released EmbeddingGemma 2, an open-weight multimodal embedding model that processes text, images, video, and audio into a single vector space. This model, with 740M parameters, is optimized for on-device AI applications like search and RAG, reducing latency and memory. It also supports Google AI Edge Foresight for Mac, enhancing context-aware meeting assistance.