Characterizing the atmospheres on planets around other stars has become a reality in the past few years. The frontier of current and near-future efforts is observing sub-Neptune-sized planets with thick atmospheres. I will present new atmosphere scenarios for GJ 1214b, the most observed mini Neptune exoplanet to date, resulted from my self-consistent atmosphere chemistry model that can treat both H2-dominated atmospheres and non-H2-dominated atmospheres on low-mass exoplanets. In addition to the conventionally assumed H2O-dominated atmosphere or hazy atmosphere, I will show that a H2-H2O atmosphere, a H2-CO atmosphere, a CO-CH4 atmosphere, and a C2H4-C2H2 atmosphere are also plausible scenarios that are consistent with the current observation of the planet. I will discuss how future observation in both transmission and thermal emission can distinguish these scenarios, and generalize the results to outline the spectral features useful for characterizing mini Neptune exoplanets. For the future when thin atmospheres on terrestrial exoplanets become observable, my atmosphere chemistry model will be pivotal as the interface between the fundamental unknowns (e.g., geological sources, biological sources, mixing rates, and escape rates) and the observables (e.g., abundances of trace gases and their spectral signatures). I will show two examples of using the atmosphere chemistry model to assess potential biosignature gas candidates: H2S as a failed biosignature gas due to its short photochemical lifetime in virtually any types of thin atmospheres, and NH3 as a biosignature gas on a planet with a N2-H2 atmosphere.