# Why Doesn’t ChatGPT Warn You Before Wasting 30 Minutes of Compute? — retained record

> Retained reddit record from reddit:ChatGPT.

Canonical: https://highsignal.app/data/records/61f7b6d9f664b3ca

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:ChatGPT)  published 2026-10-03 09:06 ingested 2026-10-04 02:35

[original source ↗](https://www.reddit.com/r/ChatGPT/comments/1wwin3p/why_doesnt_chatgpt_warn_you_before_wasting_30/)

## retained text



```
I've been using GPT-6 Astra to create 70+ page documents from 100,000+ characters of source material, and it's jump in processing is great. Here's the weird part, when I give it a massive task it works for 30-40 minutes burning through a ton of compute, only to hit me with a 5-hour usage limit error without outputting a single word. So I start a new chat, slightly shrink the task, and make it do the same thing again. 

I've had to repeat this multiple times. ChatGPT and Gemini estimate these massive runs could cost somewhere around $3-$5 each in API-equivalent compute, although obviously we don't know OpenAI's actual internal cost. If that estimate is even remotely close, wouldn't it be cheaper for OpenAI to say: "Hey, this task is probably going to exceed your usage amount. Break it into three parts?" Instead, it seems to sometimes let me spend 30 minutes of compute to produce nothing, and then I immediately start over. 

If I do this 4 times to get my document, OpenAI just spent $20 in computing power on one of my projects. Why does OpenAI allow their servers to grind for 30 minutes on a task they know will breach a limit, only to dump the progress? Isn't this an astronomical waste of server money for a $20 Plus subscription?
```



document fetched 2026-10-04 00:20 · canonical https://reddit.com/r/ChatGPT/comments/1wwin3p/why_doesnt_chatgpt_warn_you_before_wasting_30
