ORT 1.26.0 Release Candidates Available For Testing

[sanaa-hamel-microsoft]

ORT 1.26.0 will be released soon, pending successful validation. Release candidate builds are available now for testing. If you encounter issues, please report them by responding in this issue.

Release branch: rel-1.26.0
Release manager: @vraspar

Python Packages

CPU: https://aiinfra.visualstudio.com/PublicPackages/_artifacts/feed/ORT-Nightly/PyPI/onnxruntime/overview/1.26.0.dev20260505004
CUDA 12: https://aiinfra.visualstudio.com/PublicPackages/_artifacts/feed/ORT-Nightly/PyPI/onnxruntime-gpu/overview/1.26.0.dev20260505003
CUDA 13: https://aiinfra.visualstudio.com/PublicPackages/_artifacts/feed/ort-cuda-13-nightly/PyPI/onnxruntime-gpu/overview/1.26.0.dev20260505003

Nuget packages

CPU:

Microsoft.ML.OnnxRuntime
Microsoft.ML.OnnxRuntime.Managed

CUDA 12

Microsoft.ML.OnnxRuntime.Gpu
Microsoft.ML.OnnxRuntime.Gpu.Linux
Microsoft.ML.OnnxRuntime.Gpu.Windows
Microsoft.ML.OnnxRuntime.Managed

[saudet]

@sanaa-hamel-microsoft Would you have any builds with ORT Training Ops enabled for CUDA? They do not seem to compile anymore on either Linux or Windows since 1.25.0. Or are they considered deprecated?