We present new stellar models based on updated physics (opacities, expanded nuclear network and mass loss rates). We compute stellar models suitable for the mean solar neighborhood, i.e. for Z=0.02 and X=0.70. The covered mass range is from 0.8 up to 125M_{sun}_ and the models are followed until the exhaustion of carbon in the core, for the more massive ones. In addition, the effective temperatures of the more massive models are corrected for the effects of stellar winds, while models with lower effective temperatures are computed using a special treatment of the equation of state (CEFF). Convective core overshooting is assumed to be moderate and is modelled with alpha_ov_=0.20. Besides the classical ingredients of stellar models, we also provide the internal structure constants needed to investigate apsidal motion and tidal evolution in close binaries. The latter constants are made public for the first time. According to the current theories of tidal evolution, the time scales for synchronization and circularization for cool stars depend -- apart from the mass, radius and effective temperature -- on the depth of the convective envelope x_bf_ and on the radius of gyration {beta}. For stars with higher effective temperatures, these dependencies are mainly incorporated in the tidal torque constant E_2_. All these parameters are steep functions of mass and time, and thus require a special numerical treatment. The new mass loss formalism produces more mass concentrated configurations than previously, especially for more massive and more evolved stellar models. As the present grid is designed mainly for the study of double-lined eclipsing binaries, the gravity-darkening exponents necessary to calculate the surface brightness distribution in rotationally and/or tidally distorted stars are computed following the method described recently by Claret (1998, Cat. <J/A+AS/133/123>), and made available for each point of every evolutionary track.