Exoplanetary science is among the fastest evolving fields of todays astronomical research. Ground-based planet-hunting surveys alongside dedicated space missions (Kepler, CoRoT) are delivering an ever-increasing number of exoplanets, now numbering at ~690, with ESAs GAIA mission planned to bring this number into the thousands. The next logical step is the characterisation of these worlds: what is their nature? Why are they as they are? The use of the HST and Spitzer Space Telescope to probe the atmospheres of transiting hot, gaseous exoplanets has demonstrated that it is possible with current technology to address this ambitious goal. The measurements have also shown the difficulty of understanding the physics and chemistry of these environments when having to rely on a limited number of observations performed on a handful of objects. To progress substantially in this field, a dedicated facility for exoplanet characterization with an optimised instrument design (detector performance, photometric stability, etc.), able to observe through time and over a broad spectral range a statistically significant number of planets, will be essential. We analyse the performances of a 1.2/1.4m space telescope for exoplanet transit spectroscopy from the visible to the mid IR, and present the SNR ratio as function of integration time and stellar magnitude/spectral type for the acquisition of spectra of planetary atmospheres in a variety of scenarios: hot, warm, and temperate planets, orbiting stars ranging in spectral type from hot F to cool M dwarfs. We include key examples of known planets (e.g. HD 189733b, Cancri 55 e) and simulations of plausible terrestrial and gaseous planets, with a variety of thermodynamical conditions. We conclude that even most challenging targets, such as super-Earths in the habitable-zone of late-type stars, are within reach of a M-class, space-based spectroscopy mission.