# I used Hindsight as a memory provider, but it performs very poorly with context and local models. Can anyone share their experience? Has anyone used Honcho or another provider?\