NVIDIA Ising is an open suite of AI models and tools that helps quantum computing teams calibrate quantum processors and accelerate quantum error correction decoding. Ising Calibration uses a vision-language model trained on data from multiple qubit modalities to interpret experimental plots and measurements, guiding an agentic calibration workflow accessible via CLI or GUI. Ising Decoding uses compact 3D convolutional neural network pre-decoders to process localized syndrome errors before passing remaining errors to a global decoder, tunable for a processor's noise characteristics, code distance and decoding rounds. This reduces quantum processor calibration time from days to hours and makes error correction decoding viable at production speeds. Model weights, training frameworks, datasets, benchmarks and deployment recipes are open source, with an NVIDIA NIM container and hosted API also available.