Discussion on OpenAI CLI usage / consumption

[tyygy001]

I’ve been using the OpenAI CLI through Dayflow as part of my workflow, and after a few days of normal usage I started running into “usage limit reached” errors, even though I’m on a Plus subscription.

This made me wonder whether Plus is actually meant to cover sustained CLI usage, or if it’s more for lighter or occasional use. I didn’t expect to hit the limit this quickly.

Just wanted to ask:

Is this a common experience when using the OpenAI CLI on Plus?

Are there any practical ways people are using it to avoid burning through the quota so fast?

[JerryZLiu]

Hmmm, interesting - originally I had done the math on the usage and calculated that it should take up about 10% of the weekly usage limits. Are you using the CLI for other stuff? can you screenshot what you're seeing in the weekly breakdown here: https://chatgpt.com/codex/settings/usage.

[tyygy001]

I hit the limit again today 😅 Looks like this has become a regular cycle for me now. I was also wondering — does processing videos recorded on a 4K display consume more tokens than ones from a 2K display?

[JerryZLiu]

Also, if you can run npx @ccusage/codex@latest and send the results here that would be helpful in debugging!

[tyygy001]

Yeah, I find this really strange too. The only possible reason I can think of is that when I first started using it, I retried a bunch of videos that had failed to analyze earlier — that might have triggered some kind of usage restrictions.Thanks!

[WashedYirgacheffe]

I also got usage limit reached in the past 3days,and there were some failed analyze. It might be the main error.

[JerryZLiu]

this is gemini though right?

[WashedYirgacheffe]

this is gemini though right?

YES，SAME PROBLEM IN GEMINI

[JerryZLiu]

hmmm, i think there might be some simple mitigation i can do here to improve rate limiting issues. will look into this!