Finite element method is much more versatile for complex geometries. With similar meshes you should obtain however roughly the same answer for simple geometries. If not, the difference lies in the respective constitutive models and the way elasto-plasticty is handled (return mapping algorithm  etc).

I worked with FLAC and PLAXIS for solving geotechnical engineering problems. If I only compare them as tools for FDM and FEM respectively, FLAC is definitely better in terms of better convergence, if the time step is carefully selected, whereas PLAXIS has many convergence issues. The explicit time marching scheme of FLAC makes it really powerful. For liquefaction related problems, for problems involving large rock masses and large strain problems, FLAC works better. PLAXIS works better for problems with complex geometries like Bernhard A Schrefler mentioned in his answer.

I am agree with Madhavi and Bernhard. But in order to FLAC abilities in programing (I mean FISH Language) you can modify or customize your behavioral model, so it could be considered as a great and useful tool. 

I am agree with the other colleagues. I work with both codes for dynamic problems and I add to the discussion that FLAC is better for dynamic analyses: you can easily model a compliant base and adsorbing lateral boundaries. Moreover a simple non linear model working under cyclic loading condition (i.e. hysteretic damping model) is available in the library. The only shortcoming is the time step  that is very small (especially is some Rayleigh damping is added) so you suffer for large computation times for large models.

How much is it large, the model i mean that may FLAC suffer in calculations?

It is depending on the performance of your computer, the size of the elements and the duration of the motion you apply. Approximately for a 30 second input accelerogram, and a model of 300x100 m (for example a large dam resting on foundation rock), mean size of elements=2m), a 2D FLAC dynamic analyses takes about 10 hours on a medium performance PC. 

please help

Is Medium density polythene sheet is elastoplastic or elastic material in Plaxis?

i,m facing the problem of finding the ultimate tensile strength Np (kN/m) used in Geogrid properties in Plaxis. but i have the value of tensile strength in kN/m2 from the laboratory test. please let me know how this will be convert?

The core difference is the integration scheme plaxis is implicit and flac explicit