It's crazy to me that the $20 plan is so abandoned, you can hardly use it as a normal chatbot, let alone work with Code or Cowork for more than one prompt. Is Anthropic really in this huge bottleneck of compute or have they made 0 effort in trying to make their models slightly efficient??
It got worst these couple of months but lately it is pretty insane, I can confidently say that the free Codex tier gets you the same use, and you can still use ChatGPT once you run out of usage.
My question is the following. I have been thinking of making and upgrade to the $100 plan in order to start using AI in much more demanding coding tasks, while also being able to use AI in regular daily life. I really like Claude and their models, but the usage murders the experience for me. Is the $100 plan the huge difference that it should be? Do you keep bumping into the limits regularly with intense use?
For someone who has tried both worlds in the respective $100 plans, Codex o Claude. Thanks!