Why did we open-source our inference engine? Read the post

google/embeddinggemma-300m

Model Page: EmbeddingGemma

Overview

Architecture
Gemma 3
Parameters
303M
Tasks
Encode
Outputs
Dense
Dimensions
Dense: 768
Max Sequence Length
2,048 tokens
License
gemma

Benchmarks

CosQA

technology retrieval en

Code search with natural language queries

Corpus: 6,267 Queries: 500
Quality
map at 10 0.3065
mrr at 10 0.3090
ndcg at 10 0.3987
Reference →

FiQA2018

finance retrieval en

Financial opinion mining and question answering

Corpus: 57,599 Queries: 648
Quality
map at 10 0.1998
mrr at 10 0.3224
ndcg at 10 0.2635
Reference →

NFCorpus

medical retrieval en

Biomedical literature search from NutritionFacts.org

Corpus: 3,593 Queries: 323
Quality
ndcg at 10 0.3876
map at 10 0.1471
mrr at 10 0.5895
Performance RTX-4090 b1 c16
Corpus 79.6K tok/s
Corpus p50 55.7ms
Query 1.9K tok/s
Query p50 27.8ms
Reference →

SCIDOCS

scientific retrieval en

Citation prediction, document classification, and recommendation for scientific papers

Corpus: 25,656 Queries: 1,000
Quality
map at 10 0.0597
mrr at 10 0.1957
ndcg at 10 0.1091
Reference →

StackOverflowQA

technology retrieval en

Programming question answering from Stack Overflow

Corpus: 19,931 Queries: 1,994
Quality
map at 10 0.6746
mrr at 10 0.6746
ndcg at 10 0.6963
Reference →

Self-hosted inference for search & document processing

Cut API costs by 50x, boost quality with 85+ SOTA models, and keep your data in your own cloud.

Github 2.0K

Contact us

Tell us about your use case and we'll get back to you shortly.