Host openai models locally - token limit

API
1

Is it possible to host openai models on my own gpu server?
If so, is it possible to get rid of the token limit?

thanks

2

Not very practical

https://www.perplexity.ai/?s=u&uuid=f68f7720-12c7-4c07-a263-50a500d789ca