O4 mini, the best model, but I feel it more slow than yesterday

Hey team!

Has anyone else noticed o4‑mini getting really slow lately? I’ve been hitting the “We’ve reached the OpenAI limit” message and even had to plug in my own API key just to keep using it. Feels like, with all this heavy usage, they might bump it into a Premium model soon, right? :thinking:
Especially since it’s crazy cheap and unlimited now—if they keep it that way, it’d be absolute paradise for us devs (I mean, it even outperforms Claude 3.5!). What’s everyone else’s experience? Any thoughts on whether they’ll limit it or keep it free?

3 Likes

Yeah I feel the same, today it has been noticeably slower than usual, maybe they’ve made some change in anticipation of it being moved to a premium model tomorrow? Cost update for GPT-4.1 and o4-mini