How to improve voice recognition

Hello im slowly migrating across to HA and love the GUI and the simplicity of creating automation. My automation in my house is highly dependant on voice commands.I dont use microphones or speakers just my phone. I have an android phone and understand wakewords are not yet supported so to give commands i open the HA app and issue commands. Its really hit and miss, it captures around 10% of my commands. Ive tried renaming devices giving them aliases with no significant differences. Ive made some recommended chages in whisper and piper but nothing seems to improve it. The config works because on the off chance it understands. Question is my combination i.e using my phone an acceptable method. I cant remember the last time ok google failed for me.Besides sticking with google are there any alternatives? Thanks in advance.

Have you tried changing the model that Whisper uses through the addon configuration? The bigger the model, the better the recognition, however more resources are required and it may take longer to respond.

Yes at the moment its set on tint-int8 and I tried increasing beam size from 1 to 5, to be honest I don’t see any improvements. Before I try to dig deeper does voice recognition work reliable? I read that you can link google voice commands with HA but that will require a SSL domain using something like duckdns. If doing this will this defeat even using HA?
Thanks again