Model reference · open weights

embeddinggemma

embeddinggemma is an open-weight embedding model from litert-community.

For self-hosting, compare this build's requirements below with the GPU you plan to rent.

• EmbeddingGemma is a sentence-similarity model published by litert-community under the name litert-community/embeddinggemma-300m.
• It is designed for deployment on Android and iOS devices using LiteRT or the Google AI Edge RAG Library.
• The model uses the Gemma license and is distributed as a 300m variant suitable for edge computing tasks.

Summary of the litert-community/embeddinggemma-300m model card, 2026-10-01

Hugging Face ↗