
    A schema to contain the pricing information for OpenAI APIs.
    The first level field should be the name category of API, (things like "Text Tokens", "Audio Tokens", "Fine Tuning", etc.)
    The value of the first level field should be a dictionary with the following fields:
    - "pricing_table_schema": A schema for the pricing table. That is, the list of the columns
    - "pricing_table": A list of dicts, corresponding to the table of pricing information for the API category. 
        For example, ```[{"Model": ..., "Input": ..., ...,}, {"Model": ..., "Input": ..., ...,}, ...]```
        Note that Sometimes there is more than one model name, so we should also have, in there, a field for "Alternate Model Names"
    - "extras": A dictionary with extra information that might be parsed out of each category. 
    For example, there's often some text that gives a description of the API category, and/or specifies "Price per 1M tokens" etc.
        
    Note that some of these tables have a "Batch API" version too. 
    In this case, there should be extra fields that have the same name as the fields above, but with "- Batch" appended to the name.
    For example, "Text Tokens - Batch", "Fine Tuning - Batch", etc.

    It is important to note: These first level fields are not determined in advance. 
    They are determined by the data that is scraped from the page.
    Therefore, these names should not be in the schema. 
    What should be in the schema is the fact that the data should be a JSON whose first 
    level field describes a category, and whose value specifies information about these category.
    