I am maintaining a client for the OpenAI API that includes cost estimation and prices for all models, and I am confused every time a couple of new models are introduced and prices are changed. 
Recently new versions of the GPT models with reduced pricing have been announced again, but I do not understand how much the legacy models cost right now. The pricing page only lists some of the latest models. The models documentation is more comprehensive but does not list prices.
Concretely, I wonder about the pricing of:
- gpt-3.5-turbo-1106
- gpt-3.5-turbo-0613
- gpt-3.5-turbo-0613-16k
Have their prices been reduced analogously to gpt-3.5-turbo-0125 ($0.0005/1K input tokens + $0.0015/1K output tokens)? Or is it still the old prices (gpt-3.5-turbo-1106: $0.001/1K input tokens + $0.002/1K output tokens, gpt-3.5-turbo-0613: $0.0015/1K input tokens + $0.002/1K output tokens, gpt-3.5-turbo-0613-16k: $0.003/1K input tokens + $0.004/1K output tokens) and I can only look up them in the web archive? What is the general policy here?
Any pointers to the official documentation for the pricing of all legacy models would be appreciated. Obviously, if anyone from the responsible teams reads this, adding this info to the pricing page would be great. Or ideally, just provide this through a general Models API. 