Predicting the pressure distribution on an airfoil is significantly influenced by the coordinates, flight state, and the geometric shape of the airfoil. Existing methods of geometric shape representation commonly require pre-modeled airfoil data or images, which lack flexibility and overlook the potential relationship with the flight state. To address this challenge, we propose a novel Geometric-Encoded and Feature-Fused Model (GEFF), which achieves airfoil geometric encoding and feature fusion within an end-to-end model and applies it to the prediction of pressure distribution. In the model, considering the positional relationships of samples along the airfoil, we propose a novel distance attention mechanism where physical distance relationships are integrated into the calculations to generate encoded geometric features from coordinates. These encoded features and flight state features are then fused to predict the corresponding pressure distribution. Extensive ablation studies and comparison experiments were performed on GEFF. GEFF demonstrates high accuracy. Compared to Transformer, GEFF reduces the predicted mean square error by an average of 16.0%.