Announcement
Google DeepMind announces EmbeddingGemma 2 for on-device multimodal embeddings
Google says EmbeddingGemma 2 is its first natively multimodal open model for on-device embeddings, built on Gemma 4 to map text, code, images, audio, and video into a shared space locally.
TLDR
Google DeepMind announced EmbeddingGemma 2, which Google says is its first natively multimodal open model for on-device embeddings. Google says the Gemma 4-based model can map text, code, images, audio, and video into one shared space for local search and retrieval, but the specs and performance claims in the announcement come from Google’s own posts.
Combined views
1.8M
15 Sources, first seen ago
