Jina Embeddings
Embedding models designed around long inputs and multilingual coverage.
Aperçu
Jina Embeddings are embedding models built with long-context inputs as a design goal rather than a limitation to work around, so a whole document section can be embedded instead of a truncated fragment. It solves the chunking problem: when the encoder caps at a few hundred tokens, long documents must be split, and the split loses the context that made the passage meaningful. It exists because retrieval quality is often limited by what had to be discarded before embedding rather than by the model itself.
Graphe des technologies
Comment Jina Embeddings se rattache au reste de la pile.
Appliqué à
- AI Search application Applications Jina Embeddings used for AI Search Medium ✓ source
- RAG technique Techniques Jina Embeddings used for RAG High ✓ source
Alternatives
-
BGE
model
Une alternative à Jina Embeddings · source
-
Cohere Embed
model
Une alternative à Jina Embeddings · source
-
E5
model
Une alternative à Jina Embeddings · source
-
GTE
model
Une alternative à Jina Embeddings · source
-
Instructor
model
Une alternative à Jina Embeddings · source
-
Nomic Embed
model
Une alternative à Jina Embeddings · source
-
OpenAI Text Embedding 3
model
Une alternative à Jina Embeddings · source
Comparaisons
Comparaisons détaillées avec les alternatives ci-dessus.