Issues with CDFT analytical gradients

Hi, I am currently working on a script for conducting molecular dynamics with CDFT. I came across an issue with CDFT analytical gradients. Despite them being available, they provide nonsense values. Numerical gradients work fine, however slow down the simulations significantly. Has anyone else come across the issue? Have you found any solutions? I’m afraid this might be a bug in Q-CHEM itself.

Thank you for reporting. Could you please send example input and output files to Q-Chem Support (support@q-chem.com), so we can verify and investigate? Thank you.

Is this for CDFT_POP = BECKE or = FBH? I think I’ve seen a report of problems with gradients for the latter. If so, does BECKE agree with the analytic gradient for JOBTYPE = FORCE?