3D critical phenomena are in general difficult problems. In the past ??decade, powerful techniques were developed to study criticality ??following the bootstrap idea: based on consistency conditions, one ?can ??directly constrain the macroscopic theory without reference to ??microscopic details. In some cases, the constrains almost uniquely ??determine the theory and one obtains very precise critical exponent. ??In this talk, I will explain how conformal bootstrap works and ??overview recent progress. I will discuss in more detail on the 3D ?O(3) ??CFT bootstrap and its implication for cubic model.