I just do not understand how anyone could run Fable 5 regularly and find that an affordable option over hiring an actual developer...even on "medium" thinking it is insanely expensive.
@nicdunz Loved it at first. Stopped loving it as much when the false positives via the guardrails start popping up and once I realized running multiple agents runs between $200-400 an hour in extra credits.
Fable 5 is the most incredible model I have ever used. The cost is insane, and it would be worth every penny if not for anthropic blocking ~25% of my prompts. That is a dealbreaker for me. Back to Codex.
@iruletheworldmo 4.8 is better than 4.7, but I still find 5.5 better. Claude just wants to ask me so many questions. Codex knows exactly what I need and gets it done.
I am absolutely amazed every single time Codex does an automated task on my desktop while I'm away and then sends a screenshot to my phone! @OpenA@OpenAIDevs
@nicdunz I honestly don't know why I only use projects. I work in a project, even just for my agentic work flows, not even really working on a traditional coding project per se. Guess it is just an old habit.
@OfficialLoganK@Uwaynis@mercor_ai I’m definitely not saying it’s bad, because I do not think it is, but it definitely seems to apply my rules in antigravity very different than Opus and GPT do. Gemini definitely seems to take more liberties and do things that I don’t necessarily want it to do.