GTE Large
Awaiting activationAn established English embedding model for a deliberately configured serving deployment.
What it does
GTE Large is part of the General Text Embeddings family published by the original thenlper team. Its model card describes text representation for tasks such as retrieval, similarity and classification. This entry uses the original English GTE Large model.
It returns 1,024-dimensional vectors and has a short 512-token source context. embeddings.app retains it for an explicitly configured Databricks serving deployment, so an existing application can use the shared interface with the model it already depends on.
Where it helps
Existing English indexes
Keep query vectors aligned with documents already embedded using GTE Large.
Focused passage search
Retrieve related short English sections or descriptions.
Deployment continuity
Connect an intentionally configured legacy serving endpoint to one API.
When to choose it
Consider it mainly for an existing deployment or index using this exact model. For a new hosted search project, compare currently available BGE or OpenAI options first.
Awaiting a configured deployment and verified account pricing. This is the original English model, not a newer multilingual GTE variant. Keep passages short and avoid mixing vectors from different generations.
Official sources
Paraphrased from the provider’s documentation and original model information. Selection advice is our editorial guidance.
Original GTE Large model card ↗Reviewed October 3, 2026