What is EmbeddingGemma 2?

EmbeddingGemma 2 is Google DeepMind's open, lightweight multimodal embedding model that maps text, code, images, video, and audio into a single shared embedding space.

Is EmbeddingGemma 2 open source?

The model weights are released under the Apache 2.0 license and are available on Hugging Face and Kaggle.

Can EmbeddingGemma 2 run on-device?

Yes. At 740M parameters it is designed for on-device use; the full multimodal model requires roughly 567MB of RAM on a Pixel 11 Pro, and text-only use about 191MB.

Google EmbeddingGemma 2

An open, lightweight multimodal embedding model

Grands Modèles de Langage

Développement

Visiter

Informations de la fiche

Fiche d’annuaire HiCyou

Répertorié

Site web répertorié: blog.google
Fiche d’annuaire créée: 13 oct. 2026
Dernière mise à jour de la fiche: 13 oct. 2026

Informations incluses dans la fiche

Présentation
4 fonctionnalités clés
3 cas d’usage

Cette page décrit la fiche d’annuaire HiCyou. Elle ne constitue pas une certification de sécurité, de propriété ou de qualité du produit.

Vérifiez le site externe avant de partager des informations sensibles ou d’effectuer un achat.

Qu'est-ce que Google EmbeddingGemma 2

EmbeddingGemma 2 is Google DeepMind's open, lightweight multimodal embedding model. It maps text, code, images, video, and audio into a single shared embedding space, enabling cross-modal search from one query. At 740M parameters it is designed to run on-device, with weights released under the Apache 2.0 license on Hugging Face and Kaggle.

Fonctionnalités clés

Multimodal embeddings across text, code, images, video, and audio in one shared space

Lightweight 740M-parameter model that can run on-device

Open weights under the Apache 2.0 license, available on Hugging Face and Kaggle

Optimized for privacy-first use cases

Cas d'utilisation

Cross-modal semantic search across text, images, video, and audio
On-device applications requiring local, privacy-preserving embeddings
Retrieval and search features built on open embedding models

Pourquoi les startups ont-elles besoin de cet outil ?

Startups can integrate an open, Apache 2.0-licensed multimodal embedding model without heavy infrastructure, enabling privacy-first, on-device search and retrieval features across multiple media types.