AI Potluck
Contribute
Back to Gap Map Model components / Embeddings & retrieval models

EmbeddingGemma

Google
open weights / Overall score: 3.3

EmbeddingGemma is Google's 308M-parameter multilingual text embedding model, built on Gemma 3 and aimed at running on phones, laptops and tablets rather than in a datacenter. It takes up to 2K tokens of context and emits 768-dimension vectors, truncatable to 512, 256 or 128 via Matryoshka Representation Learning, and is distributed alongside a quantization-aware-trained checkpoint for low-memory deployment. The weights are downloadable but gated behind acceptance of the Gemma Terms of Use.

Openness

3 high confidence
3.0
weights
open(downloadable but gated: the Hub repo requires manual acceptance of the Gemma terms)
data
closed(approximately 320B tokens of web documents, code and synthetic text, characterized by type only)
code
partial(inference and fine-tuning guides
license
Gemma-License(Gemma Terms of Use, use bounded)

The Gemma Terms of Use permit commercial use but require every downstream copy to carry the same prohibited-use policy, a real bound almost no user hits; the click-through acceptance gate is friction, not a further restriction. The training corpus is described only by category, not published, which keeps the score at open weights.

  • https://ai.google.dev/gemma/docs/embeddinggemma/model_card recorded 2026-09-11

    Google's model card gives the 2K input context, 768-dimension output with MRL truncation, and a training dataset of approximately 320 billion tokens described by component type - web documents, code, synthetic - with no dataset released.

  • https://ai.google.dev/gemma/terms recorded 2026-09-11

    The Gemma Terms of Use grant a non-exclusive, royalty-free right to use and distribute the model, subject to the Prohibited Use Policy which must be passed on to every downstream recipient, and reserve Google's right to restrict use it believes violates those terms.

  • https://huggingface.co/api/models/google/embeddinggemma-300m recorded 2026-09-11

    gated: manual; private: false; cardData license gemma - the weights are downloadable after accepting the terms.

Adoption

4 high confidence
4.0

Adoption is measured as Hugging Face downloads across the two shipped checkpoints, almost all for embeddinggemma-300m rather than the QAT variant. The volume is notable because the repository is gated, so every download passed through a click-through license.

Capability

3 high confidence
3.0

EmbeddingGemma is strong for a 308M-parameter model and the best on-device multilingual option, but capability here reflects retrieval quality rather than efficiency, and its 61.15 on MMTEB v2 is well behind Harrier-OSS's 74.3 on the same benchmark. Its 2K context window is also a sixteenth of Harrier-OSS's, ruling out the long-document retrieval frontier models are built for.

Verified 2026-09-11