EmbeddingGemma 2: Unified and Efficient Multimodal Representations
Why I kept it
I kept EmbeddingGemma 2 because it maps text, images, audio, and video into one space at 744 million parameters, with quantized weights for on-device use. Embedding models I can actually serve have moved more slowly than large language models. BGE-M3, from BAAI (Beijing Academy of Artificial Intelligence), has been public since 2024 and stayed a strong choice for me: it covers dense, sparse, and multi-vector retrieval in more than 100 languages. Adapting a large language model for embedding search often means several billion parameters, which is hard to serve on device. Qwen3-Embedding-8B has strong performance at 8B parameters and needs more infrastructure to serve.
François Paupier
arXiv27 October 2026Public source
Read the paper
In this reference
An open embedding model that maps text, images, audio, video, and combinations of them into one vector space. At 744 million parameters, it can load modality encoders separately and includes quantized weights for resource-constrained and on-device use.
Cite the original work
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 & Omar Sanseviero & Chandramouli Amarnath & Ian Ballantyne & Zhe Li & Daniel Salz & Ibrahim Alabdulmohsin & Ilya Philippov & Kristen Quan & Glenn Cameron & Edouard Yvinec & Luciano Martins & Doug Reid & Sindhu Raghuram Panyam & Weiyi Wang & Hengxuan Ying & Madhuri Shanbhogue & Tanmaya Dabral & Shanfeng Zhang & Juyeong Ji & Renjie Wu & Joe Zou & Xu Chen & Yunan Zhang & Yu-Hui Chen & Matthew Soulanille & Vinod Mamillapalli & Yi-Chun Kuo & Wai Hon Law & Matt Kreileder & Alice Zheng & Suril Shah & Mogan Shieh & Pooja Consul & Alberto Montes & Andreas Hess & Andres Sevilla & Babak Samari & Brendan McCloskey & Brian Potetz & Daniel Cer & Georgios Kathareios & Gustavo Hernández Ábrego & Hesen Zhang & Ji Ma & Jiaxing Wu & Keegan Mosley & Paul Suganthan & Rui Meng & Shen Gao & Shengcao Cao & Yi-Ting Chen & Yichang Chen & Zhongli Ding & Kendra Byrne & Cormac Brick & Olivier Lacombe & Qin Yin & Tris Warkentin & Armand Joulin & Tom Duerig & Mojtaba Seyedhosseini (27 October 2026). EmbeddingGemma 2: Unified and Efficient Multimodal Representations [Research paper]. arXiv. https://arxiv.org/abs/2610.37364v1
View BibTeX
@misc{embeddinggemma_2,
author = {Henrique Schechter Vera and Sahil Dua and Jovan Andonov and Sara Smoot and Sebastian Russo and Ivan Llanos and Max Matveev and Jae Hyeon Yoo and Kat Black and Shih-Cheng Huang and Josiah Aklilu and Maarten Grootendorst and Omar Sanseviero and Chandramouli Amarnath and Ian Ballantyne and Zhe Li and Daniel Salz and Ibrahim Alabdulmohsin and Ilya Philippov and Kristen Quan and Glenn Cameron and Edouard Yvinec and Luciano Martins and Doug Reid and Sindhu Raghuram Panyam and Weiyi Wang and Hengxuan Ying and Madhuri Shanbhogue and Tanmaya Dabral and Shanfeng Zhang and Juyeong Ji and Renjie Wu and Joe Zou and Xu Chen and Yunan Zhang and Yu-Hui Chen and Matthew Soulanille and Vinod Mamillapalli and Yi-Chun Kuo and Wai Hon Law and Matt Kreileder and Alice Zheng and Suril Shah and Mogan Shieh and Pooja Consul and Alberto Montes and Andreas Hess and Andres Sevilla and Babak Samari and Brendan McCloskey and Brian Potetz and Daniel Cer and Georgios Kathareios and Gustavo Hernández Ábrego and Hesen Zhang and Ji Ma and Jiaxing Wu and Keegan Mosley and Paul Suganthan and Rui Meng and Shen Gao and Shengcao Cao and Yi-Ting Chen and Yichang Chen and Zhongli Ding and Kendra Byrne and Cormac Brick and Olivier Lacombe and Qin Yin and Tris Warkentin and Armand Joulin and Tom Duerig and Mojtaba Seyedhosseini},
title = {{EmbeddingGemma 2: Unified and Efficient Multimodal Representations}},
howpublished = {arXiv},
url = {https://arxiv.org/abs/2610.37364v1},
year = {2026},
version = {arXiv v1}
}
