Gemini Embedding 2：来自Gemini的原生多模态嵌入模型

Gemini Embedding 2: A Native Multimodal Embedding Model from Gemini

May 26, 2026

作者: Madhuri Shanbhogue, Zhe Li, Shanfeng Zhang, Gustavo Hernández Ábrego, Shih-Cheng Huang, Aashi Jain, Daniel Salz, Sonam Goenka, Chaitra Hegde, Ji Ma, Feiyang Chen, Jiaxing Wu, Tanmaya Dabral, Babak Samari, Kevin Poulet, Daniel Cer, Kaifeng Chen, Paul Suganathan, Hui Hui, Jovan Andonov, Philippe Schlattner, Jay Han, Iftekhar Naim, Wing Lowe, Vladimir Pchelin, Albert Yang, Yi-Ting Chen, Zhongli Ding, Grace Zhang, Georg Heigold, Yichang Chen, Antoine Reveillon, Brendan Mccloskey, Wenlei Zhou, Dahun Kim, Rui Meng, Emma Wang, Jack Zheng, Halley Fede, Zhen Yang, Keegan Mosley, Brian Potetz, Sahil Dua, Henrique Schechter Vera, Shen Gao, Hesen Zhang, Andreas Hess, Hengxuan Ying, Alberto Montes, Karan Gill, Min Choi, Sebastian Russo, Anja Hauth, Jinhyuk Lee, Michael Boratko, Megan Barnes, Vikram Rao, Claudiu Musat, Cyril Allauzen, Ehsan Variani, Shankar Kumar, Tom Bagby, Junyi Jiao, Yang Gu, Tengxin Li, Ayush Agrawal, Roberto Santana, Dev Nath, Stephen Karukas, Shuoxuan Han, Lucia Loher, Alice Twu, Nidhi Vyas, Siddharth Bhai, Frank Palma Gomez, Wangyuan Zhang, Chaoren Liu, Jizheng Yang, Steve Qiu, Shijie Zhang, Sujay Kulkarni, Sascha Rothe, Sean Nakamoto, Raphael Hoffmann, Zach Gleicher, Yunhsuan Sung, Qin Yin, Tom Duerig, Mojtaba Seyedhosseini

cs.AI

摘要

我们隆重推出 Gemini Embedding 2，一款原生多模态嵌入模型，支持将视频、音频、图像和文本等多种模态嵌入到统一的表示空间中。我们借助 Gemini 的多模态能力，为所有这些模态的任意交错输入组合生成嵌入，从而在各类任务中实现出色的泛化性能。通过在多任务、多阶段的训练框架中应用大规模对比学习，我们在关键嵌入基准测试中取得了领先水平，涵盖单模态、跨模态及多模态检索等多样化的任务。实验结果表明，我们的嵌入模型在各类任务上表现优异（在 MSCOCO 上 R@1 达 62.9，Vatex 上 NDCG@10 达 68.8，MTEB 多语言任务上达 69.9，MTEB 代码任务上达 84.0），超越了专门设计的模型。这些统一能力使 Gemini Embedding 2 成为检索增强生成（RAG）、推荐和搜索等下游应用的有力候选方案。此外，其在从天文学、生物科学到美术及烹饪艺术等不同领域的强大零样本性能，使其即使是针对专业领域，也能作为一种高度可靠、开箱即用的表示形式。

English

We introduce Gemini Embedding 2, a native multimodal embedding model that allows embedding video, audio, image, and text modalities in a unified representation space. We leverage the multimodal capabilities of Gemini to produce embeddings for arbitrary combinations of interleaved inputs across all these modalities that generalize well across a wide variety of tasks. Applying large-scale contrastive learning in a multi-task multi-stage training setup, we achieve state-of-the-art performance on key embedding benchmarks including unimodal, cross-modal, and multimodal retrieval spanning a diverse set of tasks. We show that our embedding model demonstrates strong performance (with a score of 62.9 R@1 on MSCOCO, 68.8 NDCG@10 on Vatex, 69.9 on MTEB multilingual and 84.0 on MTEB Code) across a variety of tasks surpassing the performance of specialized models. These unified capabilities make Gemini Embedding 2 a promising candidate for downstream use cases such as RAG, recommendation and search. Furthermore, its robust zero-shot performance across distinct fields - from astronomy and bioscience to fine arts and the culinary arts - establishes it as a highly reliable, out-of-the-box representation even for specialized domains.