Nora provides convenient multi-model AI services using a pay-as-you-go credit system. To help you use Nora more cost-effectively, we're sharing a simple optimization tip: Start a new chat session for different tasks or when conversations get too long.
Why Does Starting a New Chat Save Credits?
Large Language Models (LLMs) like ChatGPT, Claude, etc., need to review and process the entire chat history within the current session to understand the context when handling your requests. This means that as the conversation grows longer, the amount of information the model needs to process with each new message also increases.
Think of it like this:
The Key to Cost Optimization: Reducing Unnecessary Context Processing
When you ask a new question in a very long conversation that isn't strongly related to the earlier parts, the model still processes the entire lengthy history. This computational overhead is often unnecessary and costs you credits.
Therefore, proactively starting a new chat gives the model a "clean slate." It allows the model to focus solely on the current task without processing irrelevant historical information, effectively lowering the credit cost per request.
Recommended Practices:
By developing the habit of starting new chats when appropriate, you can gain more granular control over your credit spending, making your Nora experience more economical and efficient.