Abstract: We present results of the modelling of the Hα spectroscopic observations of two active filaments using our 2D non-LTE filament model. It is assumed in the model that filaments of such type are composed of flux-tubes located in the transition region and/or corona and relatively cool plasma can flow along these flux-tubes with various velocities. The flux-tube system is approximated in the model by a 2D horizontal slab where its finite dimensions form its cross section and the infinite dimension is parallel to the solar surface. The isothermal and isobaric slab is irradiated from the bottom and the sides and the non-LTE radiative transfer in the 2D geometry is solved using the MALI numerical technique. Assuming that flux-tubes are oriented along the magnetic field, the orientation of plasma flow in the slab is defined by the azimuth and inclination angles of the magnetic field. Moreover, all unresolved plasma motions in the slab are characterized by the velocity of micro-turbulence as an additional input parameter of the model. The first studied filament was observed on 29 May 2017 with the IBIS interferometer located at the DST telescope. It was in state of activation, and erupted next day. Only Hα spectroscopic observations of this filament were available, thus, only line-of-sight velocities of plasma flows were obtained from the modelling. Despite of this, the results clearly showed that the filament is not activated in the same way along its whole spine – some parts are hotter while other stay cool but behave very dynamic (large macroscopic velocities and high micro-turbulence). The other filament was a small-scale arch filament located in the active region AR12597 observed on 28 September 2016 with the Echelle spectrograph at the VTT telescope. Simultaneously with the Hα observations the full-Stokes spectro-polarimetric data in the He I infrared triplet were also obtained with the GRIS polarimeter located at the GREGOR telescope. These spectro-polarimetric data provided us with the possibility to estimate the magnetic field vector using the HAZEL inversion code. Then, azimuth and inclination angles of the plasma-flow velocity can be fixed in our non-LTE filament model and the absolute magnitude of the velocity can be obtained. We further plan to continue with developing our model by introducing temperature and pressure variations and fine structure (a so-called multi-slab).