Aims. X-Ray Flashes (XRFs) are soft gamma-ray bursts whose nature is not clear. Their soft spectrum can be due to cosmological effects (high redshift), an off-axis view of the jet or can be intrinsic to the source. We use SWIFT observations to investigate different scenarios proposed to explain their origin.
Methods. We have made a systematic analysis of the afterglows of XRFs with known redshift observed by SWIFT. We derive their redshift and luminosity distributions, and compare their properties with a sample of normal GRBs observed by the same instrument.
Results. The high distance hypothesis is ruled out by the redshift distribution of our sample of XRFs, indicating that, at least for our sample, the off-axis and sub-energetic hypotheses are preferred. Of course, this does not exclude that some XRFs without known redshift could be at high distance. However we find that taking into account the sensitivity of the BAT instrument, XRFs cannot be detected by SWIFT beyond approximate to 3. The luminosity distribution of XRF afterglows is similar to the GRB one. This would rule out most off-axis models, but for the homogeneous jet model. However this model predicts a GRB rate uncomfortably near the observed rate of supernovae. This implies that XRFs, at least those of our sample, are intrinsically soft.