I've been using GPT 5 via the API today. I can't get anywhere near the speed and accuracy of GPT 4.1 for our use cases. :( Thank god for the mighty Qwen3 Coder!

(This is for coding stuff which I want to be super-fast and follow instructions really closely -- GPT 4.1 is pretty good at that.)

@Jachian22 Well we have our own inference servers and custom config for qwen, so it's really the cost of that hardware I guess.