Hugging Face has released a new tool to help developers quickly determine whether their machine learning models can run on AWS Trainium accelerators without modification. The tool, called torch-neuronx model support Space, builds on TorchNeuron, a native PyTorch backend that enables Transformers models to run directly on Trainium hardware using the same code path as other devices. The compatibility checker provides a searchable interface where developers can look up popular models from the Hugging Face Hub to see whether they run unmodified on Trainium, along with performance metrics such as sharding information, compilation time, and peak device memory. The Space includes detailed instructions for running compatible models, including the exact software stack and Python scripts needed, eliminating guesswork for teams evaluating AWS's specialized hardware. Backing the tool is a public dataset with detailed compatibility results for each model tested, including error logs and version information. Hugging Face currently focuses on text-generation and image-text-to-text models, with plans to expand coverage to diffusers, sentence transformers, and other tasks in the future.