Hi all,
I’m raising this as a formal issue and strong complaint regarding the file upload system in ChatGPT, particularly for video files. I’m a free-tier user, but that shouldn’t mean I’m locked out of basic functionality that’s supposedly available to everyone.
I’ve been working with GPT-4o to analyze short sim racing footage (Assetto Corsa) — a genuinely impressive use case. However, the upload pipeline is completely broken when it comes to video:
- I’ve uploaded files in
.avi,.mp4, both under the 512MB limit. - I converted formats, re-exported, and even trimmed the videos.
- Every time, GPT-4o attempts to process the file but fails due to a timeout or internal error.
- The assistant has been helpful and patient, but it’s clear the issue lies in the backend infrastructure — not user error.
To make matters worse, as a free user I can’t submit formal feedback through the in-app tools — meaning there’s zero accountability or visibility unless I post here.
If video support is not truly functional, it should be made explicitly clear in the interface — not silently fail and waste user time.
I urge OpenAI’s devs to:
- Investigate and patch the timeout limits or decoding bugs that are breaking video input support.
- Expand upload diagnostics — if a file is unsupported or too long/complex, say so before wasting the user’s time.
- Allow free users to submit technical feedback when legitimate bugs are encountered. Locking this behind a paywall is not a good look.
This is not a “support ticket” — this is an appeal for transparency and quality. GPT-4o is incredible, but these broken features ruin the user experience and cut off valuable, creative use cases.
Thanks,
Abhijit (a very frustrated user who just wants to analyze race footage)