📷 Searching for images from text with EmbeddingGemma 2
EmbeddingGemma 2 enables on-device multimodal semantic search across photos and videos offline
“This all happens directly on my phone without any intermediate transcription, and no external API calls are required.”
Google's EmbeddingGemma 2 is a lightweight multimodal embedding model that enables natural language search across local media libraries entirely on-device, with no API calls or transcription required. Demonstrated via the Google AI Edge Gallery app, it supports both image search and video moment retrieval. The offline capability is the key differentiator, positioning it as a privacy-preserving alternative to cloud-dependent search.