High token consumption with codex and 5.5

Hi. Guys this should be fixed asap. I used it for like 5 mins and its gone. I was in gpt-5.4 high fast