Context : Circumstellar disks are expected to be the birthplaces of planets . The potential for forming one or more planets of various masses is essentially driven by the initial mass of the disks , a crucial parameter for any planet formation theory . Constraining the masses of disks is a question of great interest for low-mass stars , which are expected to harbor less massive disks . Aims : We present and analyze Herschel /PACS observations of disk-bearing M-type stars that belong to the young \sim 2 Myr old Chamaleon-I star forming region , to better constrain the properties of the circumstellar material and the stellar mass dependance of these parameters . Methods : We used the radiative transfer code RADMC to successfully model the spectral energy distributions ( SEDs ) of 17 M-type stars detected at PACS wavelengths . Our modeling strategy is carefully designed so that we search for the most probable disks parameters amongst a large grid of models , via Bayesian inference , an approach that has already proven to be successful . Results : Based on the modeling results , we first discuss the relatively low detection rates of M5 and later spectral type stars with respect to the PACS sensitivity , and argue their disks masses , or flaring indices , are likely to be low ( M _ { \mathrm { disk } } \sim 10 ^ { -5 } M _ { \odot } , \gamma \sim 1.1 ) . For M0 to M3 stars , we find a relatively broad range of disk masses ( 10 ^ { -4 } – 10 ^ { -3 } M _ { \odot } ) , scale heights , and flaring indices . Via a parametrization of dust stratification , we can reproduce the peak fluxes of the 10 \mu m emission feature observed with Spitzer /IRS , and find that disks around M-type stars may display signs of dust sedimentation . We discuss a tentative correlation between the strength of the 10 \mu m emission feature and the parametrized stratification . Conclusions : The Herschel /PACS observations of low-mass stars in Cha-I provide new constraints on their disk properties , overall suggesting that disk parameters for early M-type stars are comparable to those for more massive stars ( e.g. , comparable scale height and flaring angles ) . However , regions of the disks emitting at about 100 \mu m may still be in the optically thick regime , preventing direct determination of disk masses . Thus the modeled disk masses should be considered as lower limits . Still , we are able to extend the wavelength coverage of SED models and start characterizing effects such as dust sedimentation , an effort leading the way towards ALMA observations of these low-mass stars .