EmbeddingGemma 2 running locally in-browser on WebGPU
Summary
Google's EmbeddingGemma 2 embedding model is now running locally in the browser via WebGPU, with a Hugging Face model page and an interactive web demo showcasing on-device inference.
Similar Articles
Gemma 4 E2B running in-browser at 255 tok/s using WebGPU kernels written by Fable 5
Gemma 4 is demonstrated running in-browser via WebGPU at 255 tokens per second, using kernels generated by Fable 5, showcasing efficient on-device inference.
EmbeddingGemma 2: an open, lightweight multimodal embedding model
Google DeepMind launches EmbeddingGemma 2, a 740M-parameter open (Apache 2.0) multimodal embedding model that unifies text, code, images, video, and audio into a shared embedding space optimized for on-device inference, built on the Gemma 4 architecture with 8K context and Matryoshka truncation for storage efficiency.
google/embeddinggemma-2 · Hugging Face
Google DeepMind released EmbeddingGemma 2, an open multimodal embedding model (740M params) that maps text, images, video, and audio into a unified 768-d vector space, designed for on-device search, RAG, and classification with Matryoshka truncation for reduced storage costs.
@Saboo_Shubham_: This is Google’s Embedding Gemma 2 100% free and local running on Mac.
Google's Embedding Gemma 2 embedding model can now run 100% free and locally on Mac, enabling private and cost-free text embedding workflows.
EmbeddingGemma 2
Simon Willison comments on Google's release of EmbeddingGemma 2 under the Apache 2.0 license, arguing that open weights are essential for embedding models so users can avoid being locked into proprietary hosted services that could later be discontinued.