Research
EmbeddingGemma 2 is an open multimodal embedding model for text, images, audio and video
The model maps text, images, audio, video and mixed inputs into one shared space and is released with quantized weights, truncatable dimensions and adjustable vision token budgets. The paper reports it outperforms specialist models and models two to twelve times larger on embedding benchmarks.
Every outlet that ran it 1 article
Hugging Face PapersResearchLead
EmbeddingGemma 2: Unified and Efficient Multimodal Representations