The Kimura 2-parameter (K2P) Model is a Markov model of base substitution. It assumes the root distribution vectors describe all bases occurring uniformly in the ancestral sequence. It allows different probabilities of transitions and transversions. This means that the rate of base changes A-C and A-T are the same (transversions), and the rate of base change A-G can differ from the other two (transitions).
The transition matrix has the form $$\begin{pmatrix} \alpha&\gamma&\beta&\beta\\ \gamma&\alpha&\beta&\beta\\ \beta&\beta&\alpha&\gamma\\ \beta&\beta&\gamma&\alpha \end{pmatrix}$$