Which HF pricing plan to choose

Hello, I was wondering which HF pricing plan should I choose to back a large-scale chatbot, aka, use HF as a service provider to get the LLM needed in the chatbot.

My end goal is to use the Llama2 model without a limited rate, with low latency, and allow it to handle concurrent users.

Could you help me figure it out, please?

Thank you in advance.