Codex UI extremly slow, freezing for minutes at a time

I am not OpenAI staff, but this sounds worth splitting into two cases: model latency vs local UI freeze.

The details that would make the report easier to act on are: Codex app version and macOS version; model/reasoning setting; whether it happens in a fresh conversation or only this longer one; whether terminal commands keep running while the UI is frozen; whether CPU/memory spikes during the freeze; and one exact timestamp with timezone.

As a workaround, I would try starting a fresh conversation or trimming the long thread/context. If it still reproduces, screenshots/logs plus the delayed I am finished now state should be strong signal for support.