Google DeepMindは、テキスト、画像、音声、および動画を共有された埋め込み空間で統合したオープンソースのマルチモーダル埋め込みモデルであるEmbeddingGemma 2をリリースしました。このモデルは740億のパラメータを持ち、Gemma 4アーキテクチャに基づいて構築され、Apache 2.0ライセンスで公開されています。テキストのみのワークロードでは191MBのRAMでデバイス上での推論をサポートしており、8Kトークンのコンテキストウィンドウを備えています。
なぜ重要か. このリリースでは、プライバシーを保ちながら低遅延でマルチモーダルな検索と取得を消費者向けハードウェア上で実現し、オフラインRAGパイプラインやエッジアプリケーションを構築する開発者にとっての利点を提供しています。
Create real-time decision engines leveraging multimodal context for classification, routing, and predictive capabilities via the MediaPipe Decision Task API .
delivering a significant 9.92-point improvement on code performance (in MTEB Code, from 68.76 to 78.68)
Use your favorite development tools : Serve the model efficiently using transformers, sentence-transformers, MLX , vLLM, llama.cpp , SGLang, Ollama , and LMStudio
When paired with generative models such as Gemma 4, EmbeddingGemma 2 enables on-device RAG pipelines that understand complex multimodal data.
EmbeddingGemma 2 has 740 million parameters, making it optimal for on-device inference.
Because EmbeddingGemma 2 is built on Gemma 4 and shares its text tokenizer and audio encoder, developers can run both models together in a unified pipeline with a lower combined total memory footprint.
Develop cross-platform apps with Google AI Edge MediaPipe for turnkey embedding, retrieval & decision tasks or LiteRT for custom model integration
Store your embedding vectors with Qdrant
Develop cross-platform apps with Google AI Edge MediaPipe for turnkey embedding, retrieval & decision tasks
Build for the browser with transformers.js or WebGPU
in MTEB Code, from 68.76 to 78.68
With more than 20 million downloads, builders have used it to power smarter on-device search tools and privacy-first retrieval augmented generation (RAG) pipelines.
To learn how to build on-device search and RAG systems with LiteRT, read the Google AI Edge blog post .
Today, we’re launching EmbeddingGemma 2 , expanding beyond text to unify code, images, video, and audio in a shared embedding space. Built on the Gemma 4 architecture and released under a commercially permissive Apache 2.0 license, EmbeddingGemma 2 has 740 million parameters, ma…
EmbeddingGemma 2 is the most capable model for on-device multimodal embeddings, natively mapping combinations of text, images, audio, and video into a unified embedding space.
