@ivanfioravanti So finding that oMLX is actually quite nice and the Lightning MTP option is a significant improvement over GGUF. Any preference between Qwen and Flash as a daily or are you using GLM?
Of course OpenAI is seeing growth. They basically have unlimited tokens. No 5 hour windows and they reset every time you fall below 25% on your weekly.
@kostakent@theo@laurensnys My understanding is that Anthropic prohibits their Oauth token from being used outside of CC - so my ask is basically how one can use the proxy and mix Anthropic and non Anthropic models being used together through the proxy if that means the proxy has to hold the token.