The analysis of questions and generation of answers in an LLM require computations using GPUs, which incur costs. The basis for calculating these costs is the number of tokens.
In the AI assistant, third-party LLMs are used. The plan fees include the cost of using the LLM, and the token limit is set to ensure that the LLM usage fees do not exceed the plan fees.
