Jina Embeddings

Long-context embeddings API — 8192-token context for documents, code, and multilingual search

Jina Embeddings provides state-of-the-art text embeddings via API, specialized for long documents with 8192-token context windows. Models include jina-embeddings-v3 (multilingual, 570M parameters) and jina-colbert-v2 for late-interaction retrieval. Used for semantic search, RAG, and document retrieval where long context and multilingual support matter.

Key specs

500,000,000 Embeddings API calls monthly source

as of 2026-03-27