(I switched to using local models as usage limits, api instability and the concept of paying per token stresses me out)
(I switched to using local models as usage limits, api instability and the concept of paying per token stresses me out)