Gpt-4o-transcribe audio length limits

I think this thread from last year might be a bit outdated, but might lead to some clues?

I’d also check the docs…

https://platform.openai.com/docs/models/gpt-4o-transcribe

Good luck!