Where EmbeddingGemma 2 Breaks Down in a Real RAG Pipeline

An engineering teardown of where Google DeepMind's EmbeddingGemma 2 — a 740M open-weight multimodal embedding model that runs in 567MB and claims to outperform rivals twice its size — falls short in production RAG retrieval pipelines, from chunking and modality mixing to the gap between benchmark numbers and real-world recall.