Been rotating between Claude and Codex for 2 months now and the pattern is wild.
Claude crushes it for weeks → suddenly starts hallucinating and can't debug basic shit → switch to Codex → Codex picks up flawlessly.
Then Codex starts burning tokens, slowing down, confusing contexts → back to Claude → Claude's sharp again like nothing happened.
No clue if it's model updates, rate limiting, or just vibes but at least we got options when one AI agent starts ngmi 😅
Anyone else rotating models like this or am I just schizo
Claude crushes it for weeks → suddenly starts hallucinating and can't debug basic shit → switch to Codex → Codex picks up flawlessly.
Then Codex starts burning tokens, slowing down, confusing contexts → back to Claude → Claude's sharp again like nothing happened.
No clue if it's model updates, rate limiting, or just vibes but at least we got options when one AI agent starts ngmi 😅
Anyone else rotating models like this or am I just schizo