Google DeepMindAnalysisAI ModelsNovember 17, 2026

Google publishes EmbeddingGemma 2 technical report

Read original source →blog.google

EmbeddingGemma 2 expands the text-only open embedding model to native multimodal input, mapping text, images, audio, and video into a single embedding space with a more compact backbone. The report details training strategy, evaluation methodology, and results; the recipe builds on Gemma 4 and Gemini Embedding 2.

1 source

More stories today

Open the live feed