Hey all,
I’ve been using the Pinecone Assistant API for my project, and while it’s been reliable, I’ve noticed it’s pretty expensive per interaction, especially when handling chat or uploading documents like PDFs and text files — those really add up. The cost of embedding and storing user files is surprisingly high.
Before this, I used Pinecone indexes directly, and that setup was actually much faster and significantly cheaper. It just required more manual setup around embeddings and query routing.
I haven’t used the OpenAI Assistant API yet, but I’m curious how it compares — especially around cost and ease of use. If anyone has experience with both, how do they stack up in terms of pricing and performance? Is OpenAI’s assistant more scalable?
Would love to hear your thoughts.