EmbeddingGemma 2 EmbeddingGemma 2 is a 740-million-parameter open multimodal embedding model from Google DeepMind that natively maps text, code, images, video, and audio into a single 768-dimensional vector space. Built on the Gemma 4 architecture and released under Apache 2.0, it is optimized for on-device inference, enabling private cross-modal search, RAG, and zero-shot classification on consumer hardware.