Agree with Andre.
DIsk IO will be probably the most limiting factor. As we don't know your expected workload we can do just estimations.
YOu have 4 sata disks in R10. So you have around 320 IOPS available for all VMs. These amount IOPS are available just for 100% read. Each write IO will consume two IOs because of R10 write penalty. You can host around 10 VMs In case your typical VM consumes 30 IOPS in average.
YOu have 32GB RAM which is another limiting factor. As Andre already mentioned you have to remove ~4GB consumed by hypervisor kernel and other processes. As you want 2GB per VM it means 14 VM will be max but it can be better if VMs don't need full 2GB and ballooning is leveraged.
YOu have 12 CPU cores/24 CPU threads. Again, it depends on real workload demand. Typical rule of thumb is 3 vCPU per CPU thread which means upto 72 VMs having 1vCPU. But there are other aspects like CPU RDY and CPU Co-stop indicating performance issues.
HOpe this helps.