Hi, this might be because the vtt file is large? There is a limit how much content AI models can use as Context.
If you need to process larger files, have a look at the AI models page for the context size of each model, so perhaps chosing a model with larger context size would work.
If the file is still too large the Max mode might help, just note that the Max mode costs by number of processed tokens. That would be ok if you need to process a large file, just be aware that input and output content counts as cost. This is an advanced mode and only recommended if you are well experienced with AI prompting and know how to control context size to avoid using up your requests too fast as it costs more than one request to process large content like your VTT file. (e.g. 50kb could be 12 000 tokens approx., and 1MB could be 250 000 tokens, where those count as both input and txt tokens of same amount likely as output)
Otherwise only breaking up the file into smaller files would work and be cost effective. AI can help you also with that
Please let me know if this answers your question and mark it as solved if it helped.