Then you could try Metas ESMFold which is way faster. Also I found that you can easily predict protein sequences of up to ~2000AA relatively fast with the --cpu-only flag. As long as you have computation time available.
I don't know how available HPC is in Pakistan but from what I read on an wikipedia page you have ~10 Universities with supercomputers or something. So maybe you can get access ?
iTESSOR tool... It one of the most detailed and advanced 3D protein model prediction tool... I have used it in vaccine designing projects... It gives result in about 6 days...