In plain words: EmbeddingGemma 2 turns text, images, audio, and video into one shared set of numbers, trained in stages that gradually add modalities and harder goals. It beat specialist models and models two to twelve times larger across every modality.
Abstract
We present EmbeddingGemma 2, an open multimodal embedding model which natively represents text, images, audio, video, and combinations thereof in a unified space. Our approach introduces improvements such as simplified per-layer embeddings and a multi-stage curriculum with modality and objective progression which ensure strong quality across unimodal, cross-modal, and mixed-modal evaluations. Notably, in comprehensive embedding benchmarks across all modalities, EmbeddingGemma 2 establishes a new Pareto frontier, outperforming both specialist models and models two to twelve times larger. Furthermore, our design offers flexibility for resource-constrained environments, supporting selective modality encoder loading, adjustable vision token budgets, truncatable embedding dimensions, and quantized weights. We present ablations analyzing key decisions and release EmbeddingGemma 2 to promote further research.
Henrique Schechter Vera, Sahil Dua, Jovan Andonov, Sara Smoot, Sebastian Russo, Ivan Llanos, Max Matveev, Jae Hyeon Yoo, Kat Black, Shih-Cheng Huang, Josiah Aklilu, Maarten Grootendorst, et al.
arXiv:2610.37364 · cs.LG, cs.CL, cs.CV, cs.SD · submitted Oct 27, 2026
abstract · pdf · html · 28 pages. Models are available on HuggingFace (at https://huggingface.co/google/embeddinggemma-2) and LiteRT (at https://huggingface.co/litert-community/embeddinggemma-2-740m-litert-lm)