In the landau theory we assume order parameter that is equal to zero at $T>T_c$ and none zero at $T<T_c$ wich is valid only for order to disorder phase transition according to my understanding. So that is mean that I can't use Landau theory on Liquid - Gas phase transition?
Away from the critical point, the liquid-gas phase transition is a first order phase transition, so you must use a theoretical description that treats it as one. The standard formulation of the Landau theory does not treat first order phase transitions, but it can be modified so that it does.
The liquid-gas phase transition through the critical point is second order, and the Landau theory can be applied to it.
The Landau paradigm is getting a bit outdated in my opinion (not only) . There are plenty of different phase transitions which cannot really be categorised as first or second order. Enough to look at lattice models of quantum gravity or the standard model, but in analytic models also the study of Kibir-Żurek mechanism shows that. Many phases such as transition to spin glass, cannot be categorised according to order/disorder... Other aspects like long range entanglement can come into the picture.
The Landau description is a model related to the 20th century , when with 20th century technology we were able to create 20th century materials. Most of them could be analysed fairly easily in the labs. Nowadays we need to trap ions with lasers, excite composite systems, create a 2d lattice to give rise to anyons, create well designed lattices. One should look at phase transitions from a different perspective, because the old categorisation can easily break in many cases.
The Landau theory is not only valid to order-disorder but this is the most typical type of distinction, when your order parameter is zero or nonzero on the two sides.