We need speaker that support two way audio, so we can use it only for talking to someone inside home remotely through the internet similar to go2tc two way
You would need a microphone for recording the sound.
A speaker only plays the sound.
Some smart speakers have a microphone, but the access to it is controlled by the firmware on those speakers, which probably limit the options severally, due to privacy concerns from the vendors.
ESPHome can probably easily be made into smart speakers and there are probably also communication servers available that HA can interface with either local or cloud based.
so I can use cheap Chinese RTSP ONVIF cam with go2rtc two way audio, to use voice only 
well, yeah.
Cameras are often made to be able to receive and transmit audio along with pictures.
Speakers are rarely made to be used as two way communication.
Often a two way speaker will require echo cancellation to filter out the sound from the speaker to be any good and vendors can’t be bothered with this feature compared to how few that actually use it.