I'm running Foundry Local 1.2.1 core and SDK and the new .10 CLI release. I have tried the older CLI version which loads the CPU and GPU models but not NPU. QNNExecutionProvider loads and the Foundry log file does not include any errors.
Snapdragon(R) X2 Elite Extreme - X2E94100 - Qualcomm(R) Hexagon(TM) NPU
30.0.220.11010
2026-01-26
I read through the various other closed issues that are similar and tried everything they included to no avail. At this point I am stuck.

I'm running Foundry Local 1.2.1 core and SDK and the new .10 CLI release. I have tried the older CLI version which loads the CPU and GPU models but not NPU. QNNExecutionProvider loads and the Foundry log file does not include any errors.
Snapdragon(R) X2 Elite Extreme - X2E94100 - Qualcomm(R) Hexagon(TM) NPU
30.0.220.11010
2026-01-26
I read through the various other closed issues that are similar and tried everything they included to no avail. At this point I am stuck.
Many thanks for the update @PantsBandit. On our side we need to bump the Foundry Local SDK version in the CLI to 1.2.1, which has an enhanced fallback methodology that would have resolved your issue. I am going to update the title of the issue to align with the eastus issue and keep it open until we bump to 1.2.1.
Related to #757