Google Releases EmbeddingGemma 2, a Lightweight Sub-1B Multimodal Embedding Model Sparking Community Innovation

Google DeepMind unveiled EmbeddingGemma 2 on October 6, 2026, a 740-million-parameter open-weight model that unifies text, code, images, video, and audio into a single 768-dimensional vector space, enabling on-device multimodal search with minimal memory footprint. Released under the Apache 2.0 license, the model maps five distinct modalities into shared embeddings that fit within ~567MB RAM on consumer hardware like the Google Pixel 11 Pro, with modular configuration options ranging from 270M parameters for text-only to the full 740M stack.