Spectral cubes are now usual data products in astronomy, independently of the spectral domain, in part due to the high data production rate of large projects such as MUSE in optical, LOFAR, ALMA, VLA, NOEMA or ASKAP in radio astronomy, or Chandra and XMM in Xray. And this is only a hint of what will happen with the emergence of SKA or other Petascale projects in a near future. These cubes are, at least in part, accessible online and easy to find due to the large number of VO services which distribute them. However, efficient display and analysis of such spectral cubes is a big challenge. In this context the CASSIS team at IRAP and the Aladin team at CDS decided to work together on the combination of their VO applications in order to create a tool able to explore simultaneously both the spatial and spectral dimensions of data cubes. CASSIS is a Java tool able to discover spectra in remote services via the SSAP protocol and analyse them. It provides functionalities such as spectrum display, spectral line identification, prediction of spectra from any telescope, comparison of spectra with various models and determination of the physical parameters of the sources. Aladin Desktop is also a Java tool enabling anyone to discover and display on-sky data (images, cubes and catalogs). Images and cubes can be either in standard FITS format or in the IVOA HiPS format. Different overlays, including catalogs, survey coverages (e.g. MOC) or fields of view are available as graphic layers in this Virtual Observatory landscape that favors the comparison of data. The focus demo shows how the two desktop applications are interfaced to cooperate on data access in order to provide spectral cube analysis functionalities. For example CASSIS can analyse a spectrum built on the fly by Aladin, out of a spectral data cube, after hand selection of a specific area. Reversely CASSIS allows to select spectral ranges on such spectra and ask Aladin to display 2D images combining the corresponding spectral planes in the cube. The tool can work both on local data available on the user’s disk or on cubes discovered in VO services via the VO registry. The goal is that this new tool will favor the joined spectral and spatial exploration of data cubes in a generic way.