The article presents a Preliminary Pragmatic Model (PPM) with the practical levels of pragmatic features. Poetry translation pose great difficulty in retaining the ‘extra-linguistic features’ of the source text and target text. When placing more emphasis on linguistic features, ‘extra-linguistics’ are ignored by the evaluators. This lack is largely due to the non authentication of pragmatics of the languages involved and the non-development of ‘extra-linguistic’ parameters for evaluation. Focusing on this fact, the study takes pragmatic as the common denominator between poetry and translation, and develops a pragmatic model. The model adopts the concept and objective from the prominent Translation Quality Assessment (TQA) model of Juliane House, and the two models of Hossein Vahid Dastjerdi and his team: (a) Practical Model for Translation Analysis and Assessment of Poetic Discourse, and (b) Semiotic Model for Poetry Translation. The model is an extended part of Vahid’s model with pragmatic (features) layering. The article is prefaced with a brief description of the pragmatic features and functions to explain how and why it matters to compare poetry translation. Later, the effectiveness of the model is tested with a randomly-selected short poem by Gulzar and its English translation by Nirupama Dutt. Gulzar is an Indian poet writing in Hindustani (a mix of Hindi and Urdu). Versing mostly in free style, he makes his poems pragmatically poignant with unusual imagery. Evaluating his poems through the model verifies its working. The findings (tabulated, compared, and discussed) show literal translation of metaphors that becomes displaced as well as losing the sense in the poem. Thus, the translation misses the pragmatic force of Gulzar’s philosophy and does not achieve dynamic equivalence. The study recommends the application of pragmatic procedure for evaluation of poetry translation and offer new possibilities in translation research.