The problem is that Anthropic is overextended. Their uptime is actually a disgrace. To fix it they're attacking demand instead of supply, reducing the usage limits of the Max plan and banning secondary usage like OpenClaw.
It'll be interesting to see what happens next. OpenAI are even offering credit incentives on their enterprise subs if you sign up a Codex seat.
No one really talks about them any more in my circles. They haven't had a successful pre training run since Ilya left. The board drama was an early sign of the internal dysfunction. They lost most of their best researchers.
Also, Sam Altman is not a good person.
aurareturn•1h ago
This is because it's very easy to ask an AI to write a script to maximize tokens from your flat rate plan.