This work presents the new DassHydro computational platform which includes a sequential coupling of the SMASH (Spatially Spatially distributed Modelling and ASsimilation for Hydrology models) hydrological model and the DassFlow (Data assimilation for free surface flows) multi-D hydraulic model. The complete code has advanced Variational Data Assimilation (VDA) capabilities, applicable to the whole chain. Preprocessing tools for automatic domain meshing and coupling as well as visualisation tools are included. Numerical simulations of floods at high resolution with the integration of multi-sensor data reducing uncertainties are thus possible. Each model is built from similar technologies: efficient Fortran solvers (with computations in parallel if required), adjoint codes generated by the automatic differentiation tool TAPENADE, Python interfacing based on the F90Wrap tool. The test cases show the capabilities of information transfer from the local generation of runoff at hydrological pixel scale on catchments to downstream river network and floodplains. Moreover, given observables especially on the hydraulic domain, the calibration of large dimension hydrological parameters is possible. The DassHydro platform is available online as an open source code including detailed documentations.