LLM Model Configs?

Would it be possible for some kind of doc or updates as new models are added to get insight into their configurations?

Like I know Grok 4.1 supports a 2M token context window, but I get tbe feeling it might be capped at 128K. Could be interesting to get the option to pay more for more context windows, longer token generation, etc, unless the LLMs are all operating at their max capacities.