Jina AI introduces jina-embeddings-v3, a 570M parameter text embedding model built on the Jina-XLM-RoBERTa architecture, optimized for multilingual and long-context retrieval tasks.

— Available via the Jina Embedding API or directly through Transformers.

Jina AI introduces jina-embeddings-v3, a 570M parameter text embedding model built on the Jina-XLM-RoBERTa architecture, optimized for multilingual and long-context retrieval tasks.

— Available via the Jina Embedding API or directly through Transformers. 

https://huggingface.co/jinaai/jina-embeddings-v3

Impressive advancement in text embedding technology for multilingual retrieval tasks