Scale retrieval with async batch embeddings across providers and vector stores

We added async batch embeddings for major providers and integrated them into most vector databases. This significantly improves throughput and reduces latency and cost for data ingestion, reindexing, and large corpus updates.

Details

Works with OpenAI, Cohere, Azure OpenAI, VoyageAI, Jina, Together, Nebius, Gemini, Hugging Face, and Mistral.
Parallelize embedding jobs and push results directly into your vector store with a consistent API.

Who this is for: Teams managing large or frequently updated knowledge bases.