I only ask because making embedding model consumption based charges seem like the norm is off to me — there are a ton of open source, self managed embeddings that you can use.
You can spin up a tool like Marqo and get a platform that handles making the embedding calls and chunking strategy as well.
I only ask because making embedding model consumption based charges seem like the norm is off to me — there are a ton of open source, self managed embeddings that you can use.
You can spin up a tool like Marqo and get a platform that handles making the embedding calls and chunking strategy as well.