High CPU usage after update.
Resolved 💬 6 comments Opened Jun 10, 2026 by gutnikgleb Closed Jun 16, 2026
💡 Likely answer: A maintainer (github-actions[bot], contributor)
responded on this thread — see the highlighted reply below.
What version of the Codex App are you using (From “About Codex” dialog)?
26.608.12217
What subscription do you have?
plus
What platform is your computer?
Microsoft Windows NT 10.0.26200.0 x64
What issue are you seeing?
Codex desktop app has been updated. After that my laptops cpu almost permanently has 70-100% usage. I'm not using the app itself. It's just open. Hiding the app window means 30% CPU usage. Opening the window instantly means 70-100%.
What steps can reproduce the bug?
Feedback ID: 019eb1a2-d680-7a20-9604-1966206e9543
What is the expected behavior?
I would like there to be no unnecessary load on the processor, as there was before the application update.
Additional information
intel i5 13500h
if it has any meaning :)
6 Comments
Potential duplicates detected. Please review them and close your issue if it is a duplicate.
Powered by Codex Action
I tried deleting the app and reinstalling it. It didn't help.
Potential duplicate doesn't exactly match my situation. My processor loads even without prompts. Yes, after killing, usage drops.
Can confirm. Mine was just sitting there doing nothing at all and thrashing my CPU. Laptop was HOT. Turned off codex. Laptop cool again.
I'd also like to add another negative point after the update. Sending a prompt with a request has become significantly delayed. Previously, when I sent a request, it would be processed immediately, and I could see the model's thoughts. However, after the app update, there was a huge delay. The request doesn't go to the model, and I see a spinning loading icon on the send button. For example, now there's a delay of about two minutes between sending and the first messages about the model's thoughts.
P.S.: My internet connection is fine.
P.S.2: The long delay only occurs on the first request in a new chat.
Can confirm. Same issue. I removed all the plugins, stripped agents.md, config.yaml file and everything right back and it would still take a while before starting to answer no. A week ago it ran brilliantly, last two days its a good incentive to try out a different llm.
After todays update, issue is no longer active.
Thank you.