# Google Gemma releases EmbeddingGemma 2 for multimodal embeddings in one space

> Original title: Credit to @xenovacom for building this demo

- Source: Google Gemma
- Published: 2026-10-29T16:54:38.000Z
- Bad Influence: https://badinfluence.news/items/n7qt7oeuu1gqvkupxn2r48fjj
- Original: https://x.com/googlegemma/status/2115849805197136321

## Summary

Google Gemma announces a demo by Xenova showing EmbeddingGemma 2, which projects text, images, video, and audio into a single embedding space. The model has 740M parameters and can run entirely in a browser on WebGPU using Transformers.js.

## Post

Credit to @xenovacom for building this demo
https://x.com/xenovacom/status/2107505769503420725

## Quoted post @xenovacom

> NEW: EmbeddingGemma 2 just released on Hugging Face. Project text, images, video, and audio into a single, unified embedding space.
> 
> At just 740M parameters, it can even run 100% locally in your browser on WebGPU with 🤗 Transformers.js!
> 
> Try it yourself 👇

https://x.com/xenovacom/status/2107505769503420725