Understanding the internal computations of artificial neural networks (ANNs) is crucial for both interpretability and model optimization. NeuralNetworkCoordinates is a Python package designed not only for visualization but also for systematically studying the internal transformations within ANNs. By computing and capturing intermediate values of scaling, translation, and activation at every neuron and layer level, this package provides an in-depth analysis of the information flow through a network. Unlike conventional visualization tools, NeuralNetworkCoordinates enables researchers to quantify the functional transformations that occur within a model, offering a transparent and structured methodology for interpretability. Leveraging TensorFlow Keras for model integration and NumPy for efficient matrix operations, NeuralNetworkCoordinates facilitates detailed inspections of neuron activations and transformations. The package is particularly well-suited for visualization when the combined input–output dimensions are limited to three, allowing for graphical representation of computations. Through a structured object-oriented approach, it enables users to dissect neural networks at multiple levels, aiding research, education, and model debugging. A case study is presented, demonstrating how the package captures intermediate neuron outputs in networks approximating mathematical functions, highlighting its practical applications in understanding ANN behavior. Future work aims to expand compatibility with larger networks, incorporate benchmark comparisons, and integrate advanced visualization techniques.