Jina Embeddings
Embedding models designed around long inputs and multilingual coverage.
Resumen
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.
Grafo de tecnologías
Cómo se conecta Jina Embeddings con el resto del stack.
Aplicado a
- AI Search application Applications Jina Embeddings used for AI Search Medium ✓ source
- RAG technique Techniques Jina Embeddings used for RAG High ✓ source
Alternativas
-
BGE
model
Una alternativa a Jina Embeddings · source
-
Cohere Embed
model
Una alternativa a Jina Embeddings · source
-
E5
model
Una alternativa a Jina Embeddings · source
-
GTE
model
Una alternativa a Jina Embeddings · source
-
Instructor
model
Una alternativa a Jina Embeddings · source
-
Nomic Embed
model
Una alternativa a Jina Embeddings · source
-
OpenAI Text Embedding 3
model
Una alternativa a Jina Embeddings · source
Comparativas
Comparativas detalladas con las alternativas anteriores.