Question Regarding Langchain's Support for VertexAI Batch Processing

[jutoken]

I reviewed the documentation concerning batch processing and release, and it seems that Langchain currently handles batching by issuing multiple parallel requests to the LLM API.

On my side, I’ve already implemented batch processing with VertexAI using a JSONL file, which allows the application to make a single API call for the entire batch.

My question is: does Langchain support this type of implementation — particularly with VertexAI — or is it limited to parallel requests only?

Thank you in advance for your clarification.

Best regards,