• I'm planning to use an auto embedding field to use it in vector search, using OpenAI models. When I send a import call with a client side batch of 10K docs (i know server side it will send just 40 each time), Typesense will be able to generate all those embeddings without issues? I'm using a Typense Cloud cluster with 4 GB RAM.

Yup, should be fine.

Nice this is impressive