All modelsGoogle
Gemini Embedding 001
Multilingual Google Gemini embeddings
This model is not in the catalog right now.
See available modelsAbout the model
Gemini Embedding 001 is Google's embeddings model from the Gemini lineup, with support for many languages and high retrieval quality. It turns text into vector representations and accurately scores the semantic similarity between a query and documents in dozens of languages, which is what semantic search and RAG need. The model handles Russian confidently and is a good fit for international products. The resulting vectors are convenient for clustering, recommendations, deduplication and text classification.
Strengths
- Text in many languages, including Russian
- High retrieval quality
- Accurate semantic similarity scoring
- Vectors for clustering and recommendations
- Suitable for RAG and semantic search
- Stable results on demanding tasks
Where to use it
- Semantic search across documents
- RAG pipelines and question-answering systems
- Clustering and topic grouping
- Recommendation systems
- Deduplication and similarity search
- Text classification based on vectors