Model Releases

Google releases Gemini Embedding 2, its first native multimodal embedding model

Original title:Last week, we made Gemini Embedding 2, our first natively multimodal embedding model, available to t...

AI summary

Last week Google officially released Gemini Embedding 2, its first native multimodal embedding model, to the public. Like a "universal translator," the model converts text, image, video and audio data into unique numerical vectors. Its core breakthrough is moving beyond keyword matching to map different modalities into the same space based on semantics, thereby understanding deep connections between content. Developers have already used the model to build video analysis tools and visual shopping assistants, enabling intelligent search by taking a photo or describing a scene. The model is now available via the Gemini API or the Gemini Enterprise Agent platform.

Key Takeaways

  • Last week Google officially released Gemini Embedding 2,its first native multimodal embedding…
  • Like a "universal translator," the model converts text,image,video and audio data into unique…
  • Its core breakthrough is moving beyond keyword matching to map different modalities into the same…
  • Developers have already used the model to build video analysis tools and visual shopping…
  • The model is now available via the Gemini API or the Gemini Enterprise Agent platform.

Quick Facts

Published
Updated
Source
X:Google AI (@GoogleAI)
Category
Model Releases
AI score
69
Read the original Read the full article at the original source
Why it matters

Google 第一个原生多模态嵌入模型,把文本、图像、视频拉到同一个向量空间,做跨模态搜索的开发者可以不用再手动打标签了,但离「无感理解」还有距离。

Related entities

Source & copyright

This article is based on curated titles and AI summaries from the AIHOT public API, with attribution and links back. Original copyright remains with the original publishers and authors; this site does not mirror full texts.

Original article · AIHOT entry

How to Cite

When citing this article, keep the title, date, and permalink, and credit Orange News as the source.

Orange News: Google releases Gemini Embedding 2, its first native multimodal embedding model. May 1, 2026. https://new.nat6.net/en/news/cmolryh3501wosll966mttaas/

View all →