Google DeepMind releases Gemini Embedding 2 for AI search
Google DeepMind released Gemini Embedding 2, a native multimodal embedding model for searching text, images, video, audio, documents, and code through one system. Available via the Gemini API and Google Cloud Vertex AI, it supports AI search, RAG, and recommendations. The model achieves strong benchmark results across text, code, video, audio, and specialized domains like science and art. With embeddings up to 3,072 dimensions, it enables universities, EdTech, and research teams to search across mixed-format content including lectures, PDFs, videos, and code repositories.
